机器学习学术速递[6.16]

AI 总结：提出可分离神经架构（SNA），结合神经逼近与张量分解，通过变分框架求解偏微分方程，实现高维问题代数级缩放，并在工程案例中取得显著加速。

链接：https://arxiv.org/abs/2606.14934

机构：Kevin T. Crofton Department of Aerospace and Ocean Engineering, Virginia Polytechnic Institute and State University（弗吉尼亚理工大学凯文·T·克罗夫顿航空航天与海洋工程系）

作者：Reza T Batley, Andrew Kichline, Sourav Saha

英文摘要：This work introduces the Separable Neural Architecture (SNA), a function representational class combining neural approximation with tensor decomposition. The SNA decouples localized coordinate functions (atoms) from global interactions governed by a sparse, low-rank interaction object. This architecture possesses a compact and smooth inductive bias well-suited for solving partial differential equations (PDEs). When viewed as a Galerkin trial space under the variational SNA (VSNA) framework, the formulation satisfies classical variational guarantees under Lax-Milgram: well-posedness, quasi-optimality, convergence, and stability. In high-dimensional spatiotemporal--parametric PDEs, the VSNA mitigates the curse of dimensionality by scaling algebraically rather than exponentially. Exploiting an entirely factorized, tensor-native alternating least squares (ALS) optimization framework reduces this cost to linear in dimension. The VSNA is validated across elliptic, hyperbolic, and parabolic systems, demonstrating close alignment with predicted algebraic and spectral scaling rates. We showcase the SNA as a "solve once, query anywhere" physical world model via two engineering case studies: a 7D parametric manufacturing simulation and an experimental thermal-to-property inversion pipeline for Inconel 718. The VSNA executes a 1,000,000-query Monte Carlo sweep in 102s on a standard laptop CPU, yielding a 150,000x speedup over a full-grid finite element baseline hosted on an NVIDIA A100 GPU. It further enables real-time generative inverse-mode reconstructions under 100ms. These results demonstrate that the SNA serves as a compact mathematical substrate for continuous parameter manifolds to enable real-time inversion, optimization loops, and rapid uncertainty propagation.

AI 总结：提出有理稀疏自编码器（RSAE），用可训练有理函数替代固定编码器激活，通过两阶段流程（初始化+微调）在多种语言模型和基线激活族上提升重构与下游行为指标，不牺牲特征可解释性。

链接：https://arxiv.org/abs/2606.14990

机构：Lehigh University（里海大学）

作者：Naiyu Yin, Yue Yu

英文摘要：Sparse autoencoders (SAEs) are standard tools for mechanistic interpretability, but current SAE families are constrained by fixed encoder nonlinearities such as ReLU, JumpReLU, and TopK. This hard-codes a particular sparsity mechanism into the model and can distort the reconstruction-versus-sparsity trade-off. We introduce the Rational Sparse Autoencoder (RSAE), which replaces the fixed encoder activation with a trainable rational function. Rational activations are flexible enough to uniformly approximate the activation primitives used by existing SAE families on compact domains (for TopK, the thresholded gate obtained after a separating top-k threshold is supplied), while also providing a richer function class for adapting to the observed pre-activation geometry. We realise this idea through a two-stage pipeline: an initialisation procedure that copies the pre-trained baseline SAE weights, plugs in rational coefficients obtained by the relaxed Remez exchange on synthetic data, and calibrates the scale parameters along with the rational coefficients; followed by a fine-tuning step under the standard sparsity-regularised reconstruction objective. Empirically, on residual-stream activations of three open-weight language models and across all three baseline activation families, the RSAE strictly improves on it after the fine-tuning step, both on reconstruction-side metrics and on downstream-behaviour metrics, without sacrificing feature-level interpretability under sparse probing. These gains are consistent across host language models, across baseline activation families, and across the full range of baseline sparsity we tested, while the upgrade itself adds only a handful of scalar parameters per autoencoder and runs in minutes on a single consumer GPU.

AI 总结：训练两层Transformer编码器对有理椭圆曲线进行秩分类（rank 0/1），精度>99%，并通过机械可解释性发现模型学到了Mestre-Nagao和启发式权重，且CLS嵌入编码了L(E,1)的对数。

链接：https://arxiv.org/abs/2606.15036

作者：Pranav Venkata Konda

英文摘要：We train a two-layer transformer encoder to classify rational elliptic curves $E/\mathbb{Q}$ of conductor $\leq 10000$ as either rank 0 or rank 1 from the first 128 normalized Frobenius traces. We achieve >99% accuracy on both classes, and accuracy is essentially unchanged on test curves with no isogeny or quadratic-twist relative in the training set. We then apply techniques from mechanistic interpretability such as attention analysis, linear probing, activation patching, logit attribution, and neuron-level circuit analysis to reverse-engineer the algorithm the (centroid in function space) model learned. We find that a sparse circuit of 20 out of 512 layer-1 MLP neurons is sufficient for rank prediction under a linear probe with an AUROC of 0.992 at plateau, implementing a push-pull detector architecture of rank-0 and rank-1 detectors with a one-sided readout. However, we notice that the model has sub-optimal readout problems indicating a mismatch in rank-order between the readout pathway and the discriminative circuit. Critically, the learned input weights of the top discriminating neuron match the Mestre-Nagao sum heuristic weights $\log(p)/(p\cdot \log{B})$ with a Spearman coefficient $r = 0.997$ and Pearson coefficient $r = 0.952$: the model has learnt a result from analytic number theory from the Frobenius trace data alone. We additionally find that all 50 independently trained models concentrate CLS attention on prime positions at 2-50$\times$ the rate of composite positions. The CLS embedding encodes $\log{L(E,1)}$ with $R^2 = 0.962\pm 0.011$ across the 50 models (after controlling for the conductor). Activation patching analysis reveals that attention weights are dissociated from causal information flow. Additionally, the 50 solutions from training are near-identical in function space (with pairwise agreement $>$98.8%) despite large weight space barriers.

AI 总结：提出YB Mixer，一种基于自由费米子和广义杨-巴克斯特结构的序列令牌混合层，利用可积系统的局部代数约束保证全局计算稳定性，并实现保范正交映射、可交换传输矩阵和谱循环生成器，以支持变长序列推理。

链接：https://arxiv.org/abs/2606.15085

机构：Independent Researcher（独立研究员）

作者：Snigdha Chandan Khilar

英文摘要：The YB Mixer is a sequence token mixing layer derived from free fermion and generalized Yang Baxter structures. It applies a core principle from integrable systems where a local algebraic constraint guarantees global computational stability. By using the Ising exchange algebra the mixer creates a free fermionic structure that acts as an exactly norm preserving orthogonal map. This algebra also produces commuting transfer matrices which allow inference to be order free and adaptable to any variable budget. To ensure the model can generalize to longer sequence lengths it uses a spectral circulant generator. This generator maintains the crucial orthogonal and commuting properties of the system. The result is a highly stable and mathematically grounded architecture for sequence processing.

AI 总结：提出受控动力学吸引子Transformer（CDAT），通过耦合混合von Mises-Fisher注意力能量与Hopfield精炼能量，并引入CANN启发的兴奋-抑制调制，实现拓扑约束的动力学系统，在图异常检测和图分类任务上达到最优性能。

链接：https://arxiv.org/abs/2606.15207

机构：Xi'an Jiaotong University（西安交通大学）； Tsinghua University（清华大学）

作者：Cheng Zhang, Minnan Luo, Zesheng Yang, Ming Li, Yong-Jin Liu, Qinghua Zheng

英文摘要：Transformer architectures have dramatically advanced representation learning and inference in deep models through self-attention mechanisms. In parallel,associative memory (AM) frameworks map representations onto energy landscapes, offering interpretable retrieval mechanisms. However, their continuous-time inference dynamics lack the biological plausibility of classical Continuous Attractor Neural Networks (CANNs). To bridge this gap, we propose Controlled Dynamics Attractor Transformer (CDAT), which couples a mixture von Mises-Fisher (Mo-vMF) attention energy with a Hopfield refinement energy, while augmenting energy descent with a CANN-inspired excitation-inhibition modulation. CDAT instantiates a topology-constrained dynamical system whose couplings encode relational structure among tokens, thereby linking attractor-style dynamics to modern energy-based attention. We further provide a constructive dissipation analysis to formally establish their controlled inference dynamics. Benefiting from these robust and structured dynamics, CDAT achieves state-of-the-art performance across multiple benchmarks in graph anomaly detection and graph classification.

AI 总结：提出Hindsight Self-Distillation (HSD)方法，通过将教师模型条件于当前训练组中的成功同伴轨迹，在失败与成功轨迹的分歧处提供密集信用信号，提升LLM在数学和代码推理任务上的性能。

链接：https://arxiv.org/abs/2606.15576

机构：Department of Electrical and Computer Engineering, George Washington University（乔治华盛顿大学电气与计算机工程系）

作者：Yu Li, Shu Hong, Tian Lan

英文摘要： Reinforcement learning from verifiable rewards assigns a single scalar to each rollout, leaving token-level credit assignment underspecified in long reasoning traces. On-policy self-distillation addresses this by letting the same model act as a teacher conditioned on privileged information, producing a dense per-token signal. But the common choice of a ground-truth answer is only an endpoint cue: on terse-answer tasks, the teacher falls silent at the intermediate positions where path-level guidance matters most. We propose Hindsight Self-Distillation (HSD), which conditions the teacher on a successful peer rollout drawn from the current training group. Such a peer is an exact sample from the success-conditioned policy, requiring no additional sampled rollouts. By providing a full successful continuation rather than only the final answer, the resulting credit signal concentrates at the divergence position between a failed rollout and a successful peer. Across Qwen3-8B and Qwen3-32B on math and code benchmarks, HSD obtains the best result against GRPO variants and on-policy distillation baselines, with the largest gains on terse-answer tasks such as AIME.

AI 总结：提出Z平面神经网络，通过有界几何激活函数Radial Bounding将隐藏状态映射到超球面上的2D相量束，在保持方向信息的同时限制能量幅度，理论证明其保持1-Lipschitz连续性并防止梯度消失，实验表明100层无ReLU和LayerNorm的MLP在MNIST上稳定收敛。

链接：https://arxiv.org/abs/2606.15669

机构：College of Pharmacy, Chungnam National University（忠南大学药学院）； Department of Computer Science & Engineering, Chungnam National University（忠南大学计算机科学与工程系）

作者：Sungwoo Goo, Hwi-yeol Yun, Sangkeun Jung

英文摘要：Modern deep neural networks rely on Euclidean scalar activations (e.g., ReLU) and global normalization techniques (e.g., LayerNorm) to prevent gradient instability in deep architectures. However, these mechanisms inherently cause dead neurons, discard critical directional information, and destroy the orthogonality of feature representations. Inspired by the frequency-modulation transmission of biological axons, we propose the Z-Plane Neural Network, which maps hidden states into 2D phasor bundles on a hypersphere. We introduce a novel geometric activation function, Radial Bounding($\mathbf{x} / \max(1, \|\mathbf{x}\|_2)$), which limits the energy magnitude while preserving the phase (direction). We demonstrate mathematically that this isotropic activation maintains 1-Lipschitz continuity and prevents gradient vanishing by preserving tangential gradients. Empirically, a 100-layer Z-Plane Multi-Layer Perceptron (MLP)-entirely devoid of ReLU and LayerNorm-successfully converges on the MNIST dataset with 98.34% accuracy and absolute numerical stability, proving that bounded geometric activation alone is sufficient for stable deep learning.

AI 总结：提出储层注意力网络（RAN），通过在预训练Transformer中间层注入固定随机储层来携带跨前向传播状态，实验表明未训练的循环动态足以传递可用状态。

链接：https://arxiv.org/abs/2606.15678

作者：Emma Leonhart

英文摘要：A feasibility and dynamics study of the Reservoir Attention Network (RAN), an architecture that injects a fixed, randomly-initialized reservoir into the mid-layer attention of a pretrained transformer to carry state across forward passes. Experiments span GPT-2 (124M, 355M) to Qwen2.5 (0.5B, 1.5B) on a single consumer GPU. The tasks are minimal probes chosen to isolate individual mechanisms; the broader always-alive agent vision is treated throughout as compute-limited future work, not a claim of this paper. The reservoir is left untrained (fixed random) by design: this isolates whether untrained recurrent dynamics alone suffice to carry usable cross-pass state, leaving trained recurrence as a complementary, more expensive direction.

AI 总结：针对一类支持向量机在大规模数据集上的可扩展性问题，提出将数据集分解为单个样本并训练单数据支持向量机，再通过集成学习组合模型，同时采用数据缩减策略加速，实验表明该方法在保持分类性能的同时显著提升速度。

链接：https://arxiv.org/abs/2606.16002

机构：University of Hradec Králové（赫拉德茨-克拉洛韦大学）； Universiti Teknologi Malaysia（马来西亚理工大学）

作者：Toshitaka Hayashi, Dalibor Cimr, Hamido Fujita, Richard Cimler

英文摘要：One-class classification (OCC) is a classification problem in which the training data contains only one class. The one-class support vector machine (OCSVM) is one of the most competitive OCC algorithms. However, OCSVM has scalability issues with large-scale datasets. This paper proposes the acceleration strategy of OCSVM. The idea is to decompose the dataset into samples and train OCSVM models for single data points. Subsequently, ensemble learning is applied to combine all models to compute the OCSVM model for the dataset. In addition, further acceleration is achieved through a data-reduction strategy with an OCSVM model trained on the average of the training samples. The experiment compared the proposal and traditional OCSVM using the Python package. The proposed strategy is faster than traditional OCSVM, while achieving similar classification results. Moreover, the proposed strategy can create one-to-one correspondence between samples and models. Source code is uploaded at this https URL

AI 总结：提出将时序分类错误分解为表征错误和决策错误，通过冻结原生分类器并添加残差多尺度分支与事后分支感知校准器，在不重训练骨干网络的情况下区分缺失时序证据与未充分利用的决策级证据。

链接：https://arxiv.org/abs/2606.16034

机构：Department of Computer Science (DCC), Universidade Federal de Minas Gerais (UFMG)（米纳斯吉拉斯联邦大学计算机科学系）

作者：Arthur Chagas, Arthur Buzelin, Yan Aquino, Pedro Bento, Gisele L. Pappa, Wagner Meira Jr., Cristiano Arbex Valle

英文摘要： Temporal classification errors are often treated as representation failures, but they can also arise from how available evidence is converted into decisions. This paper proposes a representation--calibration decomposition for temporal classification. We keep a trained native classifier frozen and separate two inference-time interventions: a conservative residual multi-scale branch that adds auxiliary logits to the native prediction, and a post-hoc branch-aware calibrator that recombines native and residual evidence at decision time. This design distinguishes missing temporal evidence from underused decision-level evidence without retraining the backbone. Across FI-2010, PTB-XL, UCI-HAR, MHEALTH, and HARTH, we find that gains are strongly regime-dependent. Residual multi-scale evidence is most useful in noisy or representation-limited settings, especially short-horizon FI-2010 and weaker recurrent backbones, while branch-aware calibration helps when native and auxiliary logits contain complementary evidence not fully exploited by the raw decision rule. Near-saturated settings show limited gains from either intervention. These results suggest that temporal classification should be understood not only as representation learning, but also as the problem of trusting, combining, and calibrating evidence from multiple views.

AI 总结：提出Phys-JEPA架构，将物理一致性约束引入潜在状态和状态转移，分解预测状态为物理和残差分量，在气候、交通、电力数据集上提升预测精度。

链接：https://arxiv.org/abs/2606.16076

机构：Tianjin University（天津大学）

作者：Weizhi Nie, Weichao Liu, Honglin Guo, Yuting Su

英文摘要：Multivariate forecasting in physical systems requires models that predict coupled temporal variables while preserving meaningful state evolution. Deep forecasters can fit temporal correlations, and physics-informed models can regularize predictions with scientific constraints, but these directions are often connected only at the decoded-output level. As a result, the hidden predictive state that generates future trajectories may remain statistically useful but physically unstructured. We introduce Phys-JEPA, a physics-informed joint-embedding predictive architecture for multivariate time-series forecasting. Phys-JEPA learns a latent world model in which predictive states are decomposed into physical and residual components, and physical consistency is imposed directly on latent states and latent transitions rather than only on decoded forecasts. This formulation uses known physical variables to organize the representation space while retaining residual capacity for unresolved dynamics. On Jena Climate 2009--2016, Phys-JEPA reduces aggregate MSE from 0.12482 to 0.12273 and temperature MSE from 0.01892 to 0.01831 at H=24. On Traffic, full Phys-JEPA improves aggregate MSE over the supervised baseline across all tested horizons, reducing H=192 MSE from 0.800784 to 0.773873. On Electricity, the best variant depends on horizon: static latent consistency is strongest at H=24 and H=48, while full Phys-JEPA gives the best aggregate and target-variable MSE at H=192. These initial results suggest that moving physics-informed learning from output space to latent predictive state space is a promising direction for interpretable temporal world models.

AI 总结：针对残差网络中残差流范数随深度增长导致深层更新被抑制的问题，提出范数无关残差架构NAG，通过分离幅度和方向信息保持各层贡献，并实现可解释的自适应深度跳过机制，在等计算量下匹配全深度性能。

链接：https://arxiv.org/abs/2606.16112

作者：Tomás Figliolia, Beren Millidge

英文摘要：Residual architectures are ubiquitous in deep learning, but they suffer from a subtle structural limitation: the norm of the residual stream can grow rapidly with depth. As a result, updates from later layers become small relative to the accumulated residual state. This reduces their impact on the representation and limits the benefits of scaling models in depth. To address this, we introduce NAG, a norm-agnostic residual architecture that separates magnitude from directional information in the residual stream, preserving meaningful layer contributions throughout depth and preventing later updates from being systematically suppressed by residual-norm growth. Importantly, NAG introduces only a negligible number of additional parameters and relies on simple operations that are easily kernel-fusible, preserving training efficiency in practice. We show that this architecture outperforms baseline Transformers, with gains that increase substantially as depth grows, enabling effective training of much deeper models. The norm-agnostic formulation also leads to an interpretable Mixture-of-Depths (MoD) mechanism that adaptively skips both attention and MLP layers. Beyond serving as a post-training accuracy-compute tradeoff, this mechanism can be used as a pretraining-time scaling strategy: under iso-FLOP training, compute saved by reducing per-token forward-pass cost can be reinvested into training on more tokens while keeping the total parameter count and KV-cache budget fixed. In our experiments, moderate Mixture-of-Depths rates of approximately 20%-25% match full-depth baseline performance under equal training compute while substantially reducing the number of executed layer parameters and forward-pass FLOPs. These results identify sparsity in depth as a new scaling axis for fixed-compute training, enabling very deep yet FLOP-efficient models.

AI 总结：提出CuSeT方法，通过自适应令牌级掩码和区域感知样本重加权，在简单SFT框架内提升LLM生成CUDA内核的功能正确性。

链接：https://arxiv.org/abs/2606.16231

机构：Shanghai Jiao Tong University（上海交通大学）； Biren Technology（壁仞科技）

作者：Wentao Chen, Jiace Zhu, Xing Zhe Chai, Zeng Qu, Qiaoling Xiao, Liucheng Duan, An Zou

英文摘要： High-performance CUDA kernels are essential for scalable AI systems, while Large Language Models (LLMs) still struggle to generate correct kernels due to strict and implicit execution constraints. Existing LLM-based approaches either rely on costly agentic or reinforcement-learning (RL) pipelines, or adopt supervised fine-tuning (SFT) objectives that fail to explicitly model CUDA sensitivity, namely code tokens or regions tightly coupled with execution constraints. In this work, we investigate CUDA sensitivity from the perspective of token confidence patterns, showing that CUDA sensitivity appears at both token and region levels, where most CUDA-sensitive tokens are predicted with high confidence, while a smaller low-confidence subset forms regions corresponding to execution-critical structures. These findings suggest that effective CUDA kernel generation should both leverage high-confidence CUDA-sensitive tokens and preserve low-confidence CUDA-sensitive regions. Building on these insights, we propose \textbf{\underline{CU}DA-\underline{Se}nsitive Instruction \underline{T}uning (CuSeT)}, a low-cost post-training method within a simple SFT framework. CuSeT follows the principle of ``from tokens to regions'' by combining \emph{adaptive token-level masking} with \emph{region-aware sample reweighting}. Experiments show that CuSeT consistently improves functional correctness across multiple model families and scales, outperforming standard SFT and advanced SFT variants, while achieving competitive performance against frontier CUDA kernel generation models with substantially lower inference cost.

AI 总结：提出基于线性规划的局部搜索框架，通过双层优化联合更新模型参数和正则化超参数，利用验证梯度和Hessian信息构造局部下降方向，在保持训练最优性的同时减少过拟合，实验表明在GPT-2 Small微调中持续改善测试困惑度。

链接：https://arxiv.org/abs/2606.16243

机构：Department of Management Sciences, Indian Institute of Technology Kanpur（印度理工学院坎普尔分校管理科学系）； Department of Civil Engineering, Indian Institute of Technology Kanpur（印度理工学院坎普尔分校土木工程系）； Operations and Decision Sciences, Indian Institute of Management Ahmedabad（印度管理学院艾哈迈达巴德分校运营与决策科学系）； Brij Disa Centre for Data Science and AI, Indian Institute of Management Ahmedabad（印度管理学院艾哈迈达巴德分校Brij Disa数据科学与人工智能中心）

作者：Abhishek Shukla, Anikeit Khanna, Ankur Sinha, Faiz Hamid

英文摘要：This paper proposes a Linear Programming (LP)-based local search framework for fine-tuning pretrained transformer models with explicit control against overfitting. The approach formulates transformer fine-tuning as a bilevel optimization-based regularization problem, in which model parameters and regularization hyperparameters are jointly updated. Information collected during initial warm-up iterations, including validation gradients and training Hessian information, is used to construct a local descent direction by solving an LP that minimizes a scaled directional derivative while preserving training optimality. This validation-aware descent direction enables focused local updates of both parameters and regularization hyperparameters, reducing overfitting without requiring repeated full retraining cycles. The resulting method, termed Linear Programming-based Fine-Tuning (LiFT) for transformers, differs from conventional fine-tuning by systematically identifying task-specific updates rather than relying on heuristic or grid-based hyperparameter selection. Experiments on GPT-2 Small fine-tuned on WikiText-2 demonstrate that LiFT enables effective adaptation through selective tuning of transformer blocks and regularization parameters, yielding consistent improvements in test perplexity across multiple layer configurations and regularization settings, with particularly pronounced gains in overfitting-prone scenarios. Beyond empirical performance, LiFT establishes a principled connection between transformer fine-tuning, bilevel optimization, local search, and regularization theory.

AI 总结：提出QK归一化与MLA兼容的方法，通过吸收静态权重和动态标量，无需缓存完整键，在400M模型训练中降低损失并提升下游精度，解码延迟增加小于2%。

链接：https://arxiv.org/abs/2606.16310

机构：The Chinese University of Hong Kong（香港中文大学）； Ant Group（蚂蚁集团）

作者：Yizhou Han, Yao Zhao, Jun Zhou, Longfei Li, Ruoyu Sun

英文摘要：Query-key (QK) normalization stabilizes attention by controlling the scale of queries and keys before the dot product, but is not immediately compatible with Multi-head Latent Attention (MLA). MLA achieves efficient decoding by caching low-dimensional latent states instead of full keys, whereas post-projection QK RMSNorm appears to require the fully projected key for every cached token. We show this apparent incompatibility is an implementation artifact, not an architectural constraint. RMSNorm decomposes into a static affine weight and a dynamic scalar RMS statistic. The static key-side weight can be absorbed into the MLA query-side projection; the dynamic key statistic reduces to one inverse-RMS scalar per token and KV group. The resulting formulation is exactly equivalent to explicit post-projection QK RMSNorm in exact arithmetic and preserves MLA's latent decode path. In our 400M runs trained for up to 100B tokens, QK-Normed MLA achieves lower training loss and better downstream accuracy than QK clipping, while H800 decode benchmarks show less than 2% latency overhead up to 256k context. These results make QK normalization a practical stabilization option for MLA models without requiring full-key caching.

AI 总结：提出CacheMuon，通过缓存历史优化步的极分解因子来减少Muon优化器中牛顿-舒尔茨迭代的计算开销，在保持训练质量的同时降低正交化计算量。

链接：https://arxiv.org/abs/2606.16371

作者：Bishnu Dev (1), Sushil Bohara (1), Martin Takáč (1), Samuel Horváth (1) ((1) Mohamed bin Zayed University of Artificial Intelligence, Abu Dhabi, UAE)

英文摘要： Muon is an optimizer that computes updates using the polar factor of the momentum matrix and has shown strong empirical performance across a range of training settings. A key component of Muon is the Newton-Schulz iteration used to compute this polar factor. Although this avoids the cost of an exact singular value decomposition, it remains expensive in practice because it is applied at every optimization step. At the same time, the momentum matrix changes smoothly over training, suggesting strong temporal correlation in the corresponding polar factors. In this paper, we exploit this structure and propose CacheMuon, a temporal preconditioning method that reuses information from previous optimization steps to approximate the polar factor at the current step. This reduces redundant orthogonalization computation across iterations. We analyze CacheMuon as an inexact Muon update, with error controlled by fresh-solver error and cache staleness. Empirically, CacheMuon provides a controllable quality-efficiency frontier: conservative thresholds closely match fresh Muon on language-model and vision training while reducing orthogonalization FLOPs, whereas more aggressive thresholds yield larger arithmetic savings at the cost of modest validation-quality degradation.

AI 总结：提出KaBiN模型，结合Kaiming初始化和偏差校正，解决高维不完全张量补全中初始化不当和偏差缺失导致的优化不稳定问题。

链接：https://arxiv.org/abs/2606.16388

机构：School of Computer Science and Engineering, Chongqing University of Science and Technology（重庆科技大学计算机科学与工程学院）； College of Computer and Information Science, School of Software, Southwest University（西南大学计算机与信息科学学院 软件学院）

作者：Yuchao Su, Yixin Ran

英文摘要：High-dimensional incomplete (HDI) tensors are widely used in traffic and climate applications, but sparse observations make accurate completion difficult. The intrinsic non-linear dynamics and non-stationary variations across distinct multi-modal fields severely hinder the efficacy of conventional linear reconstruction frameworks. Neural Tucker factorization provides an effective framework for modeling high-order interactions among tensor modes. By parameterizing underlying structural characteristics into continuous latent spaces, neural representations circumvent the rigid low-rank constraints of classical algebra. However, its performance can still be affected by implementation-level choices, especially parameter initialization and the bias configuration of the final output mapping. Suboptimal initializations frequently lead to variance explosion across the cubically expanded interaction spaces, driving the subsequent non-linear activation boundaries into severe gradient saturation zones, while the omission of a dedicated translation parameter forces interaction weights to implicitly absorb global statistical deviations. This paper proposes a simple yet effective neural Tucker factorization model with Kaiming initialization and bias correction (KaBiN) for HDI tensor completion. The proposed model utilizes Kaiming uniform initialization for the embedding and Tucker linear parameters, and adopts a simple bias correction in output mapping. By elegantly decoupling global mean shifts from local structural representations, the framework provides a highly stable and well-conditioned optimization landscape. Experiments on three real-world HDI tensor datasets show that KaBiN achieves better performance than the original NeuTucF, while introducing minimal computational overhead.

AI 总结：提出Taylor-Calibrate方法，利用泰勒引导的教师注意力统计初始化混合线性注意力学生模型，显著减少蒸馏所需训练令牌数。

链接：https://arxiv.org/abs/2606.16429

机构：The University of Sydney（悉尼大学）； Together AI； University of California, Berkeley（加州大学伯克利分校）； The University of Texas at Austin（德克萨斯大学奥斯汀分校）； Microsoft（微软）

作者：Zhongzhu Zhou, Qingyang Wu, Junxiong Wang, Mayank Mishra, Shuaiwen Leon Song, Ben Athiwaratkun, Chenfeng Xu

英文摘要：Hybrid linear attention models offer an appealing path to faster long-context inference: they reduce the quadratic cost and KV-cache burden of full softmax attention while retaining much of the quality of Transformer models. A practical way to obtain such models is to convert a pretrained Transformer instead of pretraining a new architecture from scratch, but this conversion is still brittle. Simply copying the teacher attention projections into a Gated DeltaNet (GDN) student does not specify the new recurrent decay, write, and output-gating dynamics. As a result, the converted model often starts in a poor dynamical regime and must spend many distillation tokens repairing initialization rather than learning the remaining teacher behavior. We propose Taylor-Calibrate, a lightweight initialization method for hybrid GDN students. The method uses Taylor-guided teacher attention statistics to set the value projection, memory timescale, write gates, and output gate, then applies a short per-layer alignment step to match each converted layer to the teacher output. Across four teacher settings and three retained-layer policies, Taylor-Calibrate gives substantially stronger zero-shot students, with up to an 88x improvement in a representative ablation, and reaches matched recovery targets with 4.9x--9.2x fewer training tokens than naive conversion.

AI 总结：提出SDS-LoRA，通过结构解耦奇异值与反向传播，消除LoRA中梯度各向异性缩放导致的秩降低和次优对齐问题，提升收敛速度和适应性能。

链接：https://arxiv.org/abs/2606.16454

机构：Seoul National University（首尔大学）； Hanyang University（汉阳大学）

作者：Junghun Oh, Sungyong Baik, Kyoung Mu Lee

英文摘要：Low-Rank Adaptation (LoRA) enables efficient adaptation of large pre-trained models to downstream tasks by parameterizing weight updates with low-rank matrices. In this paper, we investigate the limitations of the LoRA parameterization from a geometric perspective. Specifically, we show that when a full fine-tuning gradient is backpropagated to the low-rank matrices, it undergoes anisotropic scaling driven by their singular values. We argue that this phenomenon is undesirable because it distorts the full fine-tuning gradient by skewing it toward dominant singular directions while suppressing others. Our analyses demonstrate that anisotropic gradient scaling reduces the effective rank of the low-rank matrices' gradients and results in suboptimal alignment between the full fine-tuning gradient and its low-rank approximation in LoRA, thereby exacerbating the gap to full fine-tuning. To address these limitations, we propose a new low-rank parameterization, SDS-LoRA, which structurally decouples singular values from the backward pass. Our method ensures that the full fine-tuning gradient backpropagates only through the orthonormal bases of the low-rank matrices' subspaces, independent of their scales. Convergence analysis demonstrates that while LoRA's convergence rate degrades with the condition number of the low-rank matrices, SDS-LoRA remains independent of it. Experimental results across natural language and vision benchmarks show that SDS-LoRA improves loss convergence and reduces the gap to full fine-tuning, significantly enhancing adaptation performance.

AI 总结：提出SPRI方法，利用预训练FFN权重的SVD分解残差初始化MoE专家，结合两阶段训练策略，在数据受限的多语言语音翻译任务中显著提升性能。

链接：https://arxiv.org/abs/2606.16456

机构：Northeastern University, China（东北大学）； Huawei TSC, China（华为技术有限公司）； CUHK-Shenzhen, China（香港中文大学（深圳））； University of Maryland, USA（马里兰大学）； Harbin Engineering University, China（哈尔滨工程大学）； Inclusion AI, Ant Group（蚂蚁集团Inclusion AI）； NiuTrans Research, China（小牛翻译研究中心）

作者：Weiqiao Shan, Ruixiang Mao, Yuang Li, Yuhao Zhang, Yingfeng Luo, Tong Zheng, Chen Xu, Yucheng Qiao, Chunxiang Jin, Yi Yuan, Jingdong Chen, Tong Xiao, Jingbo Zhu

英文摘要：Mixture-of-Experts (MoE) models enable efficient scaling, but training them from scratch remains prohibitively expensive. MoE upcycling mitigates this cost by converting pretrained dense models into sparse MoE models. However, existing upcycling methods typically rely on large-scale continued training and often perform poorly under data-constrained supervised adaptation, due to either homogeneous experts or overly disruptive perturbations to pretrained parameters. In this setting, effective upcycling must leverage pretrained weight structure while introducing sufficient diversity among routed experts. To this end, we propose SVD-Partitioned Residual Initialization (SPRI), which distributes SVD-partitioned residuals derived from pretrained feed-forward network (FFN) weights across routed experts, introducing controlled expert diversity grounded in pretrained spectral structure. We further introduce a two-stage training strategy to improve adaptation stability. We evaluate SPRI on multilingual speech-to-text translation, where limited supervised data challenges MoE upcycling and multiple target languages provide natural routing heterogeneity. On CoVoST2 across 15 En-to-XX directions, SPRI improves average BLEU and COMET over fully fine-tuned dense models by 2.58 and 3.32 points, respectively, and outperforms the prior best MoE upcycling baseline by 3.39 BLEU and 4.34 COMET points.

AI 总结：针对深度神经网络在捕捉振荡和多尺度行为时的谱偏差问题，提出RepNet模型，通过重参数化第一隐藏层的权重和偏置，有效控制初始斜率尺度和分区点分布，实现自适应频率缩放，在函数逼近、PDE求解和算子学习中显著提升精度。

链接：https://arxiv.org/abs/2606.16575

机构：Institute of Interdisciplinary Research for Mathematics and Applied Science, School of Mathematics and Statistics, Huazhong University of Science and Technology（华中科技大学数学与统计学院交叉科学与应用数学研究所）； Institute of Computational Mathematics, Academy of Mathematics and Systems Science, Chinese Academy of Sciences（中国科学院数学与系统科学研究院计算数学研究所）

作者：Yong Wang, Tao Zhou, Xuhui Meng

英文摘要：Deep neural networks (DNNs) have achieved remarkable success in scientific computing, yet they often suffer from spectral bias in capturing oscillatory and multiscale behaviors. In this study, we investigate this limitation by examining the failure of shallow ReLU neural networks in fitting high-frequency functions. This observation identifies two important factors in resolving rapid oscillations: the initial slope scale and the distribution of partition points induced by the networks. Motivated by this analysis, we propose RepNet, a reparameterized DNN model for ReLU and tanh networks designed for high-frequency and multiscale problems. The key idea is to reparameterize the weights and biases in the first hidden layer, which enables effective control of the initial slope scale and provides an appropriate distribution of the initial partition points. Furthermore, treating the reparameterized weights and biases as trainable parameters allows the DNN to achieve adaptive frequency scaling during training. In addition, we derive quantitative estimates for the output and slope magnitudes of the reparameterized DNN to guide the initialization of the proposed method. Numerical experiments, including multiscale one- and four-dimensional function approximation, forward and inverse PDE problems in combination with physics-informed neural networks (PINNs), and operator learning, demonstrate that RepNet improves the predicted accuracy of vanilla DNNs in capturing highly oscillatory features with slightly additional computational cost. These results indicate that RepNet provides an effective and flexible approach for overcoming spectral bias and applying DNNs to multiscale problems.

AI 总结：提出熵门控潜在递归（EGLR），通过在高不确定性token处递归应用冻结模型顶层解码器，构建与温度采样正交的确定性采样轴，扩展推理时缩放空间，在数学推理任务中显著提升性能。

链接：https://arxiv.org/abs/2606.16620

机构：Mohamed bin Zayed University of Artificial Intelligence（穆罕默德·本·扎耶德人工智能大学）

作者：Soham Bhattacharjee, Dushyant Singh Chauhan, Salem Lahlou, Martin Takac, Nils Lukas

英文摘要： Inference-time scaling has become the dominant lever for improving language-model reasoning, but existing methods derive rollout diversity from a single source: stochastic token-level sampling. We argue that this single-axis sampling space is fundamentally limiting, and identify a second, fully deterministic and complementary axis: the layer span $L$ at which a frozen model's top decoder layers are recursively re-applied at high-uncertainty tokens. Different choices of $L$ produce distinct rollouts that solve different subsets of problems, with no stochasticity. We instantiate this axis through Entropy-Gated Latent Recursion (EGLR), a training-free decoding procedure that re-applies the top-$L$ layers for at most $K_{\max}$ iterations until the next-token distribution converges. Combined with $T$ temperature samples, EGLR turns a single-axis stochastic rollout pool into an $L\times T$ Cartesian sampling space at almost the same per-rollout cost. We characterize this space across $8$ instruction-tuned models and $6$ math reasoning benchmarks, and show that the $L$-axis is genuinely complementary to temperature: on MATH-500 with Qwen2.5-3B-Instruct, the joint $L\times T$ oracle reaches $91.6\%$, $+8.2$ percentage points beyond the temperature-only oracle ($83.4\%$) and $+10.4$ points beyond the layer-only oracle ($81.2\%$), confirming that the two axes capture genuinely complementary problems. The expanded rollout pool provides richer per-prompt candidates for any downstream procedure that consumes rollouts, including self-consistency, best-of-$N$ with verifiers, and group-relative RL training (GRPO), opening a new direction for inference-time scaling that does not rely on stochastic noise.

AI 总结：提出SPICE框架，利用部分信息分解理论动态量化样本复杂度，设计渐进式课程使模型从学习共享跨模态线索过渡到模态特定模式再到复杂协同交互，在多个多模态基准上取得一致改进。

链接：https://arxiv.org/abs/2606.16639

机构：New York Institute of Technology（纽约理工学院）； New York University（纽约大学）

作者：Ankush Pratap Singh, Houwei Cao, Yong Liu

英文摘要：Multimodal learning exploits complementary information across heterogeneous modalities. The informativeness of each modality can vary widely across samples and training stages. Existing multimodal curriculum learning strategies often assume that the relative complexity of samples remains unchanged throughout training and therefore cannot adapt to model evolution. We propose SPICE (Synergy and Partial Information based Curriculum Evolution), a novel progressive curriculum framework for multimodal interaction learning. Guided by Partial Information Decomposition (PID) theory, our approach decomposes multimodal interactions into redundant, unique, and synergistic information components, enabling an interpretable and dynamic characterization of sample complexity. Building on this decomposition, we design a progressive curriculum that evolves throughout training, allowing the model to transition from learning shared cross-modal cues to modality-specific patterns and, finally, to complex synergistic interactions. Adapting to model evolution, sample ordering is refined in real-time using PID information estimates derived from unimodal and multimodal predictions. Experiments across multiple multimodal benchmarks demonstrate consistent improvements over conventional training and state-of-the-art baselines, highlighting the effectiveness of PID information decomposition and adaptive sample ordering for multimodal curriculum learning.

AI 总结：提出深度变换器作为平均场交互系统实现分布式推理的理论，利用函数向量逐层推断潜在上下文变量，在上下文回归任务中预测非高斯分层结构与深度的关系，并通过约束线性注意力变换器验证。

链接：https://arxiv.org/abs/2606.16694

机构：Joseph Henry Laboratories of Physics, Princeton University（普林斯顿大学约瑟夫·亨利物理实验室）

作者：Ravin Raj, Gautam Reddy

英文摘要：Transformers are widely used as a general-purpose substrate for learning complex correlations between a large collection of coupled variables, but their internal mechanisms have remained mysterious. We introduce a theory of a deep transformer as a mean-field interacting system that implements distributed inference, subject to constraints on communication, locality and depth. We show that such a system can exploit internal state representations ('function vectors') to infer a latent context variable at increasingly finer scales over its layers. In an in-context regression task, the theory predicts a non-trivial relationship between non-Gaussian, hierarchical structure in the latent context variable, and transformer depth. Predictions are tested using constrained linear attention transformers and demonstrate adaptive inference in deep architectures. Feedforward blocks and depth enable transformers to implement a much richer class of in-context learning algorithms than previously described.

AI 总结：提出基于热启动幂迭代的快速在线曲率估计方法，并发现训练不稳定性与预条件曲率激增相关，进而提出渐进增加网络深度的架构预热策略，有效稳定大模型训练。

链接：https://arxiv.org/abs/2606.16768

机构：Pluralis Research

作者：Sameera Ramasinghe, Ajanthan Thalaiyasingam, Hadi Mohaghegh Dolatabadi, Chamin Hewa Koneputugodage, Gil Avraham, Violetta Shevchenko, Yan Zuo, Karol Pajak, Alexander Long

英文摘要：Training billion-parameter Transformers is often brittle, with transient loss spikes and divergence that waste compute. Even though the recently developed Edge of Stability (EoS) theory provides a powerful tool to understand and control the stability of optimization methods via the (preconditioned) curvature, these curvature-controlling methods are not popular in large-scale Transformer training due to the complexity of curvature estimation. To this end, we first introduce a fast online estimator of the largest (preconditioned) Hessian eigenvalue (i.e., curvature) based on a warm-started variant for power iteration with Hessian-vector products. We show theoretically, and verify empirically, that the proposed method makes per-iteration curvature tracking feasible at billion parameter scale while being more accurate. Using this tool, we find that training instabilities coincide with surges in preconditioned curvature and that curvature grows with depth. Motivated by these observations, we propose architecture warm-up: progressively growing network depth to carefully control the preconditioned Hessian and stabilize training. Experiments on large Transformers validate that our approach enables efficient curvature tracking and reduces instabilities compared to existing state-of-the-art stabilization techniques without slowing down convergence.

AI 总结：针对Muon等优化器在大模型预训练中增益随规模增大而减弱的问题，提出Hyperball包装器，固定权重矩阵及其更新的Frobenius范数，在1.2B参数模型上实现20-30%的token等效加速，并改善学习率迁移。

链接：https://arxiv.org/abs/2606.16899

机构： Stanford University（斯坦福大学）； Princeton University（普林斯顿大学）； Tsinghua University（清华大学）

作者：Kaiyue Wen, Xingyu Dang, Kaifeng Lyu, Tengyu Ma, Percy Liang

英文摘要：Matrix based optimizers such as Muon can substantially speed up language model pretraining, but their gains over AdamW are observed to shrink as model size and data scale grow when using standard constant decoupled weight decay. We propose Hyperball, a simple optimizer wrapper that addresses this issue. Given a base optimizer such as Adam or Muon, Hyperball sets the Frobenius norms of weight matrices and their corresponding optimizer updates to fixed constants. On Qwen3 style models up to 1.2B parameters, Muon Hyperball achieves 20--30% token equivalent speedup over weight decay baselines. Hyperball also improves learning rate transfer across widths and depths compared to decoupled weight decay. This method is motivated by prior theory showing that training with weight decay leads to an equilibrium weight norm that only depends on the training hyperparameters. Through this mechanism, the weight decay then decides the angular learning rate, i.e. how fast the direction of the weight matrix changes.

AI 总结：提出因子化神经算子（FaNO），通过分解谱表示为等变动态响应和不变持久响应，提升多尺度物理系统的预测精度、参数效率和跨尺度泛化能力。

链接：https://arxiv.org/abs/2606.16900

机构：School of Medicine, University of Dundee（邓迪大学医学院）； School of Data Science, Fudan University（复旦大学数据科学学院）； School of Mathematical Sciences, Fudan University（复旦大学数学科学学院）； Institute of Science and Technology for Brain-inspired Intelligence, Fudan University（复旦大学类脑智能科学与技术研究院）； School of Science and Engineering, University of Dundee（邓迪大学科学与工程学院）； Department of Applied Mathematics and Theoretical Physics, University of Cambridge（剑桥大学应用数学与理论物理系）

作者：Hao Tang, Yuechen Duan, Jiongyu Zhu, Zimeng Feng, Hao Li, Chao Li

英文摘要：Physical systems often exhibit heterogeneous mechanisms, where rapidly evolving dynamics coexist with persistent structures. Capturing such multiscale physical behavior remains challenging for existing neural operators, which typically rely on single dominant inductive bias and therefore couple distinct physical responses into a shared representation. We introduce the Unified Green's Function Framework across domains and propose the Factorized Neural Operators (FaNO), which decompose spectral representations into equivariant dynamic responses and invariant persistent responses, leading to better interpretability and generalization. Mechanistically, we show that the two operator branches spontaneously specialize into distinct physical roles that remain consistent across scales and domains: the equivariant branch captures rapidly varying transient dynamics, whereas the invariant branch extracts coherent persistent structures. This factorized mechanism of FaNO improves prediction accuracy, parameter efficiency and cross-scale generalization across physical systems and domains. In particular, it maintains consistent predictions under long-horizon autoregressive rollout, cross-resolution extrapolation and physical-regime shifts. These findings suggest that scalable physical modeling may benefit from moving beyond single-inductive-bias formulations toward factorized operator representations that better reflect the heterogeneous organization of physical systems, accelerating the reliable deployment of machine learning for scientific computing and discovery.

AI 总结：提出CircuitLasso方法，基于稀疏线性回归高效学习LLM中的稀疏电路，以SAE特征为单元，在保持结构准确性的同时大幅降低计算成本，并揭示语义特征传播机制。

链接：https://arxiv.org/abs/2606.16939

作者：Naiyu Yin, Dennis Wei, Tian Gao, Amit Dhurandhar, Karthikeyan Natesan Ramamurthy, Yue Yu

英文摘要：A prominent research direction in mechanistic interpretability is learning sparse circuits over LLM components to reveal how they jointly produce model behavior. However, raw neurons are polysemantic, making learned circuits hard to interpret. Sparse autoencoder (SAE) features alleviate this, but their high dimensionality makes existing intervention-based circuit learning methods computationally prohibitive. We propose CircuitLasso, a scalable circuit-learning approach based on sparse linear regression. CircuitLasso recovers circuits whose structural accuracy matches that of state-of-the-art intervention-based methods on the benchmark data, at a fraction of the computational cost. For interpretability, CircuitLasso efficiently uncovers relationships among SAE features, showing how human-interpretable semantic features propagate through the model and influence its predictions. Finally, we validate the utility of our learned circuits by leveraging their insights to achieve comparable performance at substantially lower cost on a domain-generalization task.

AI 总结：提出SPaiK方法，利用随机广义vec技巧（sGVT）实现成对核学习的大规模扩展，在七个药物-靶标亲和力数据集上优于现有方法。

链接：https://arxiv.org/abs/2606.16979

机构：Department of Computing, University of Turku（图尔库大学计算系）

作者：Napsu Karmitsa, Tapio Pahikkala, Antti Airola

英文摘要：Pairwise learning is a specialized form of supervised learning that focuses on predicting outcomes for pairs of objects. In this work, we introduce SPaiK, a new scalable kernel learning method tailored for pairwise settings. Our approach preserves the expressive power of kernel methods while substantially reducing computational and memory requirements. The key innovation is the stochastic generalized vec trick (sGVT), a stochastic extension of the sparse Kronecker product multiplication algorithm, which enables efficient large-scale training with pairwise kernels. By incorporating sGVT, SPaiK makes it possible to apply kernel-based pairwise learning to datasets of a size previously out of reach. We evaluate the performance of SPaiK on seven real-world drug-target affinity datasets and compare the results with state-of-the-art methods in pairwise learning.

AI 总结：提出HAMON无源衍射光学预测核心，通过光学传播替代数字序列混合层，在多个基准上优于或接近最强数字基线，MSE最多降低14%。

链接：https://arxiv.org/abs/2606.17028

作者：Alper Yıldırım

英文摘要：Simple linear and frequency-domain models remain surprisingly competitive in long-horizon time-series forecasting, and recent mechanistic evidence suggests that standard forecasting benchmarks may not require the dense superposed representations that make transformers powerful in other domains. This raises a substrate-level question: if the core forecasting operator is often low-complexity and approximately linear, does it need to be implemented as learned digital temporal mixing? We introduce HAMON, a passive diffractive optical forecasting core in which historical values are encoded onto an optical aperture, future positions are left dark, and cascaded trainable phase masks with free-space diffraction shape the forecast directly in the output field. At inference, prediction is performed by a single passive optical propagation pass with no trainable digital sequence-mixing layer. Across standard benchmarks, HAMON outperforms the strongest digital baselines considered on ETTm2 at all horizons and on ETTh2 at all but the longest horizon, improving MSE by up to 14\% and doing so consistently across horizons rather than at isolated points. It is competitive on Weather and trails the strongest baselines on the remaining ETT settings and on the high-channel-count Traffic and Electricity datasets. Phase encoding, intensity-compatible readout, and phase-scrambling ablations, together with a TorchOptics cross-simulator check, indicate that the forecasts arise from the data-bearing optical field rather than from a digital forecasting head. Because the passive core uses standard Fourier optics, HAMON defines a concrete target for optical hardware and for passive physical sequence mixing.

AI 总结：针对稀疏自编码器中内积评分受输入范数干扰的问题，提出余弦评分方法，使特征检测更关注方向对齐，实验表明该方法能更频繁地学习到人类可识别的概念。

链接：https://arxiv.org/abs/2606.15054

作者：Silen Naihin, Lev Stambler

英文摘要：Sparse autoencoders (SAEs) detect features via inner product, so a feature's activation scales with both its directional alignment and the input's norm. Under BatchTopK, high-norm tokens inflate all pre-activations simultaneously, claiming dictionary slots regardless of content alignment. This matters because sublayer normalization has already discarded the magnitude the score measures, so the encoder detects a quantity the model does not read. We replace the score with a learned blend of cosine similarity and input magnitude, letting the optimizer choose how much norm to use; a per-feature extension lets each feature decide independently. In both regimes, training is free to recover inner product but never does, with no feature ever choosing more than half-magnitude dependence. At matched reconstruction, the cosine encoder learns features that align with human-recognizable concepts far more often than standard, filling dictionary slots that inner product wastes on norm detectors. Loss reweighting that equalizes gradients barely closes the gap, confirming forward-pass score geometry as the lever. The advantage is not universal across tasks or depths, but we believe cosine scoring should be the default for dictionary learning on normalized representations.

AI 总结：针对现有激活引导方法仅捕捉均值差异的局限，提出无训练的高维随机投影激活引导方法（HiDRA），通过在投影高维空间中进行激活加法，捕获非线性特征子空间中的判别信号，实验证明其优于基线方法。

链接：https://arxiv.org/abs/2606.15092

机构：Hanoi University of Science and Technology（河内科技大学）； National University of Singapore（新加坡国立大学）

作者：Minh-Hieu Pham, Bach Do, Laziz Abdullaev, Tan Minh Nguyen, Khoat Than

英文摘要：Activation steering has emerged as a key methodology for controlling the behavior of large language models (LLMs). Existing difference-in-means based methods, however, are fundamentally limited: they capture only mean differences between class activations and fail to recover discriminative signals that naturally exist in the nonlinear feature subspace under the superposition hypothesis. Motivated by that, we propose High-Dimensional Random-projection for Activation Steering (HiDRA), a training-free approach that integrates seamlessly with existing activation steering methods. By performing activation addition in the projected high-dimensional space, HiDRA can provably capture a better discriminative structure beyond the reach of linear methods. Experiments across diverse LLM families and benchmarks demonstrate that HiDRA consistently outperforms baseline counterparts, achieving stronger behavioral control without significant computational overhead.

AI 总结：提出RECTOR自监督框架，通过自适应功能分区和掩码拓扑学习，统一建模EEG/sEEG的区域-通道-时间动态，在情感识别和任务参与分类上达到新最优，且对缺失通道和跨导联泛化鲁棒。

链接：https://arxiv.org/abs/2606.15278

机构：Cornell University（康奈尔大学）

作者：Jinhan Liu, Mahsa Shoaran

英文摘要： Affective and cognitive disorders manifest as distributed, time-varying brain network dynamics across regions, channels, and time, challenging robust representation learning from EEG/sEEG for clinical diagnosis. We propose RECTOR (Masked Region-Channel-Temporal Modeling), an end-to-end self-supervised framework that unifies joint region-channel-temporal representation learning beyond fixed anatomical priors. At its core, RECTOR-SA is a hierarchical, block-sparse self-attention induced by Adaptive Functional Partitioning that evolves region structures from static anatomical definitions to adaptive functional regions. The self-supervision is driven by Masked Topology and Representation Learning, which jointly optimizes three complementary objectives: Masked Predictive Modeling, Topological Structure Modeling, and Cross-View Consistency. Across diverse benchmarks, RECTOR sets a new state-of-the-art in EEG emotion recognition and sEEG task-engagement classification. Crucially, its strong robustness to missing channels and cross-montage generalization underscores its potential for large-scale pre-training on heterogeneous EEG/sEEG, providing interpretable insights at both region and channel levels.

AI 总结：提出UL4M4框架，通过无监督聚类和迭代插补处理任意缺失模态，实现跨模态结构保持和尺度不变性，在超过50%模态缺失时仍稳定达到F1-Micro>0.7。

链接：https://arxiv.org/abs/2606.15743

机构：Bournemouth University（伯恩茅斯大学）

作者：Hassan Ismkhan, Hamid Bouchahcia

英文摘要：This paper addresses the missing-modality challenge in multi-modal learning by introducing Unsupervised Learning for Missing Modalities in Multi-Modal Learning (UL4M4), a flexible framework that imputes missing feature embeddings in a task-independent manner before supervised prediction. We propose modality-specific normalization and a novel partial-modality distance metric to enable fair clustering of incomplete observations, capturing cross-modal structures while preserving scale-invariance across varying dimensionalities and modality counts. Cluster centers from this unsupervised stage guide an iterative greedy imputation process for any missing modalities during training or inference, supporting arbitrary numbers of modalities and arbitrary missing patterns per sample. The imputation module is lightweight, uses frozen encoders, and decouples from the downstream task, allowing easy integration with any fusion/prediction architecture. Extensive experiments under diverse and highly incomplete regimes demonstrate UL4M4's robustness, achieving, to the best of our knowledge, the first consistent F1-Micro scores above 0.7 on challenging missing configurations even when more than 50\% of modality slots are missing. Results are also stable across cluster sizes and significantly outperform state-of-the-art baselines. Code is available here: this https URL.

AI 总结：提出ProGenMech框架，通过跨层稀疏编码器忠实恢复ProGen3的生成计算，并零样本发现与蛋白质生成和适应性预测相关的稀疏电路，揭示生物意义基序。

链接：https://arxiv.org/abs/2606.16044

作者：Darin Tsui, William Deinzer, Daniel Saeedi, Amirali Aghazadeh

英文摘要：Protein language models (pLMs) can generate novel protein sequences with properties beyond those observed in nature, yet the mechanisms underlying protein generation remain poorly understood. Existing mechanistic interpretability methods based on sparse autoencoders and transcoders primarily focus on protein representation learning models and do not capture the computation required for autoregressive generation. Here, we introduce ProGenMech, a mechanistic interpretability framework for generative protein language models that extends cross-layer transcoders (CLTs) to ProGen3, a sparse Mixture-of-Experts model trained for both causal generation and span infilling. Unlike per-layer approaches, CLTs reconstruct each layer using sparse latent variables from all preceding layers, enabling faithful recovery of inter-layer generative computation. We further develop a zero-shot circuit discovery framework to identify sparse latent circuits responsible for protein generation and fitness prediction. In causal generation and zero-shot fitness estimation tasks, ProGenMech outperforms local transcoder baselines in recovering ProGen3's probability distribution and functional scoring behavior, while matching the original model's generative distribution in span infilling tasks. Moreover, the recovered circuits reveal biologically meaningful motifs and functional regions associated with conserved sequence patterns and protein fitness landscapes, establishing a foundation for interpretable and steerable protein generation.

AI 总结：针对现有表示相似性度量缺乏可解释性、对异常值敏感且计算复杂的问题，提出基于序数相似性的三元组和四元组相似性指数，实现可解释、鲁棒且高效的对齐度量。

链接：https://arxiv.org/abs/2606.16379

作者：Diogo Soares, Pankhil Gawade, Andrea Dittadi, Ewa Szczurek

英文摘要：Evaluating representation similarity is fundamental to representation learning. However, existing metrics suffer from significant limitations: they lack interpretability due to shifting baselines, lack robustness to outliers, and are computationally intractable for large datasets, forcing reliance on heuristic approximations. To address this, we develop an ordinal-similarity framework, instantiated by the Triplet (TSI) and Quadruplet (QSI) Similarity Indices, which measure alignment by quantifying the consistency of ordinal relationships. We theoretically demonstrate this formulation is inherently interpretable, robust to outliers, and computationally efficient. Finally, we establish a formal equivalence between TSI and local neighborhood alignment, measured by Mutual Nearest Neighbors. Empirically, we validate these properties and show that ordinal similarity offers a scalable approach to measuring alignment, enabling practitioners to better understand and design representations.

AI 总结：提出使用主体特定编码器替代共享编码器，结合共同分类器实现跨主体脑电图对齐，实验表明该方法能内化欧几里得对齐的作用，提高类别区分度，并识别出未见主体的编码器选择是主要瓶颈。

链接：https://arxiv.org/abs/2606.16462

机构：University of São Paulo（圣保罗大学）； Université Paris-Saclay, Inria TAU team, LISN-CNRS（巴黎萨克雷大学，Inria TAU团队，LISN-CNRS）； Institut de neuromodulation, GHU Paris, psychiatrie et neurosciences, centre hospitalier Sainte-Anne, pôle hospitalo-universitaire 15, Université Paris Cité（神经调控研究所，GHU巴黎，精神病学与神经科学，圣安娜医院，大学医院中心15区，巴黎西岱大学）； Federal University of ABC (UFABC)（ABC联邦大学）； Yneuro； Swartz Center for Computational Neuroscience (SCCN), Institute for Neural Computation (INC), University of California San Diego（斯沃茨计算神经科学中心，神经计算研究所，加州大学圣地亚哥分校）

作者：Bruna J. Lopes, Gabriel Schwartz, Sylvain Chevallier, Raphael Y. de Camargo, Bruno Aristimunha

英文摘要：Cross-subject EEG decoding promises more training data, but it also exposes neural networks to strong inter-subject distribution shifts. We study whether task supervision and architecture alone can learn subject-aligned representations. We replace a shared EEG encoder with subject-specific encoders followed by a common classifier, and compare this hybrid model with standard EEGNet, AttentionBaseNet, and CTNet baselines with Euclidean Alignment (EA) on four motor-imagery datasets. EA improves shared encoders by recentering subject covariances, but the hybrid encoder largely internalises this role: validation-loss curves and latent-distance analyses change little when EA is removed. Subject-specific heads increase class distinctiveness and place each subject close to its own latent manifold, improving most subjects while leaving a method-sensitive subset. These results support subject-specific encoders as a learned alignment mechanism for EEG decoding and identify head selection for unseen subjects as the remaining bottleneck.

AI 总结：提出QPILOTS方法，在推理时通过投影去噪中间状态到最终动作估计并计算评论家梯度来引导流匹配和扩散策略，无需修改原策略，在离线到在线RL基准上达到90%平均成功率。

链接：https://arxiv.org/abs/2606.14801

机构：University of Toronto（多伦多大学）； Vector Institute（向量研究所）； LG Electronics（LG电子）

作者：Yifan Ruan, Chenyang Cao, Andreas Burger, Ali Pesaranghader, Kaveh Kamali, Jaehong Kim, Nandita Vijaykumar, Alan Aspuru-Guzik, Igor Gilitschenski, Nicholas Rhinehart

英文摘要：Flow-matching and diffusion policies are expressive action generators, but optimizing them with temporal-difference reinforcement learning (RL) remains difficult. Effective policy extraction requires exploiting the critic's action gradient, yet directly backpropagating this signal through a multi-step denoising process can be numerically unstable. Existing methods work around this either by discarding gradient information, distilling the policy into a simpler one-step actor, or repeatedly fine-tuning the denoising policy as the critic improves. We propose QPILOTS, a method that leaves the original policy unmodified and steers the denoising process at inference time. At each denoising step, instead of evaluating the critic on the noisy intermediate action where critic predictions are unreliable, we first project that intermediate state to an estimate of the final clean action and compute the critic gradient there. We introduce two variants: QPILOTS-U uses a fast single-point approximation, while QPILOTS-M draws differentiable posterior samples via a learned auxiliary network. On a standard offline-to-online RL benchmark, QPILOTS achieves the best aggregate performance, reaching an average success rate of 90% across 50 tasks. We also apply QPILOTS to steer a large, frozen, pretrained Vision-Language Action (VLA) foundation model, outperforming or matching prior inference-time approaches across six manipulation tasks in simulation.

AI 总结：针对推荐系统中嵌入模型路由问题，形式化为具有低秩专家的对抗性上下文线性赌博机，提出Hypentropy策略梯度算法，实现$\tilde{\mathcal O}(s\sqrt{M T})$线性化策略遗憾。

链接：https://arxiv.org/abs/2606.14929

机构：Operations Research Center, MIT（麻省理工学院运筹学研究中心）； Sloan School of Management, MIT（麻省理工学院斯隆管理学院）； Department of EECS, MIT（麻省理工学院电气工程与计算机科学系）

作者：Yan Dai, Negin Golrezaei, Patrick Jaillet

英文摘要：Modern recommendation systems increasingly rely on dynamically routing diverse queries to multiple embedding models. Despite its practical significance, this problem remains poorly understood under realistic conditions like adversarial queries, bandit feedback, and limited observability of models. We formalize embedding model routing as an adversarial contextual linear bandit with low-rank experts, where contexts are queries, actions are items, and experts are the embedding models working on low-rank latent representation spaces. We first establish that standard regret notions suffer from structural misspecification or statistical intractability, and we identify a log-quadratic policy class that is expressive enough to capture query-dependent model routing, yet structured enough to allow efficient online learning. Second, we propose a policy gradient algorithm called Hypentropy Policy Gradient (HPG). It provably adapts to the unknown low-rank structure under incomplete information and attains $\tilde{\mathcal O}(s\sqrt{M T})$ linearized policy regret -- where $s, M$, and $T$ are the intrinsic rank of the experts, the number of models, and the number of rounds -- thus avoiding a curse of dimensionality. Finally, we also provide an computationally efficient and parameter-free implementation of HPG.

AI 总结：提出上下文多臂赌博机框架，通过学习个体溢出概率并排序连接用户，以最大化激励口碑奖励，实验证明考虑溢出异质性可提升目标定位精度。

链接：https://arxiv.org/abs/2606.15146

作者：Ahmed Sayeed Faruk, Elena Zheleva

英文摘要：Stimulated word-of-mouth is a strategy that promotes information sharing through prompts or incentives. Optimizing stimulated word-of-mouth through social networks requires identifying and targeting connected users who are most susceptible to spillover, a phenomenon where the influence of recommendations extends beyond the immediate audience to impact their connected users. The probability of spillover varies across individuals, and their connections, leading to heterogeneity. Understanding and accurately estimating the spillover probabilities among users in social networks is crucial for improving the effectiveness of stimulated word-of-mouth. To address this, we present a novel contextual multi-armed bandit framework that learns individual spillover probabilities and ranks connected users to maximize rewards from stimulated word-of-mouth. Experiments on real-world network datasets demonstrate that accounting for spillover heterogeneity enhances the targeting precision of top-$k$ connected users, boosting rewards and outperforming baseline methods that do not learn individual spillover effects.

AI 总结：提出StarOR框架，结合蒙特卡洛树搜索与测试时强化学习，通过四阶段分解和GRPO更新LoRA适配器，实现无监督细粒度奖励的中间决策优化，在5个基准上以4B模型达到最优性能。

链接：https://arxiv.org/abs/2606.15197

机构：School of Computer Science and Engineering, Southeast University（东南大学计算机科学与工程学院）； Northwest A&F University（西北农林科技大学）

作者：Jiajun Li, Yu Ding, Shisi Guan, Ran Hou, Wanyuan Wang

英文摘要：Optimization modeling is inherently hierarchical, requiring a precise sequence of symbolic commitments. Traditional learning-based automated optimization modeling methods improve modeling policies through large-scale annotated or curated training data, but are costly to adapt to new problem distributions. Meanwhile, one-shot generation remains brittle in hierarchical modeling, where early symbolic errors can propagate into invalid formulations. Test-time scaling offers a promising alternative by enabling structural exploration with additional instance-level computation; however, existing search-based methods typically rely on a fixed policy, causing repeated rollouts to inherit similar modeling biases and providing limited credit assignment for intermediate decisions. To address these limitations, we propose StarOR, a synergistic search-and-adaptation framework that couples MCTS with Test-Time Reinforcement Learning for optimization modeling. StarOR decomposes the modeling process into four stages and updates a transient LoRA adapter via GRPO at each non-terminal node. By using MCTS-generated siblings as local comparison sets, StarOR transforms search-time exploration into instance-specific policy refinement. Moreover, an unsupervised multi-faceted reward system provides fine-grained feedback for intermediate formulation decisions without ground-truth labels. Experiments across five optimization benchmarks show that StarOR achieves state-of-the-art performance even with a 4B backbone, outperforming existing methods and the frontier LLMs.

AI 总结：本文通过构造反例证明，在表格设置下，蒙特卡洛探索性初始状态（MCES）算法可能收敛到次优解，并提出基于状态级学习率缩放的修正方法以恢复最优性收敛。

链接：https://arxiv.org/abs/2606.15247

机构：Department of Engineering, University of Cambridge（剑桥大学工程系）

作者：Octave Oliviers, Glenn Vinnicombe

英文摘要：The asymptotic behaviour of Monte Carlo Exploring Starts (MCES) is a long-standing open question in reinforcement learning, even in the tabular setting. We investigated the convergence properties of tabular MCES by constructing examples in which the algorithm converges to suboptimal solutions. This paper presents new counterexamples for both initial-visit and first-visit MCES and gives a convergence-restoring modification for the initial-visit case. We show that stable suboptimal solutions may exist for initial-visit MCES with sample-average updates even when greedy actions are updated more often than non-greedy actions on average. However, by scaling learning rates inversely to update frequencies on a state-by-state basis, convergence to optimality is guaranteed. Unlike previous uniformisation methods, this modification is applicable to large-scale problems that require approximating the estimated value function. We then extend the example to show that sample-average first-visit MCES may also converge to suboptimal solutions. This largely settles a fundamental open problem and shows that exploring starts alone do not guarantee convergence to optimality. More broadly, these results highlight that convergence depends critically on the relative size and frequency of updates applied to different actions, making the choice of learning rates and the balance between exploration and exploitation central to the analysis of MCES and the implementation of scalable Monte Carlo control methods.

AI 总结：提出TruDi方法，通过信任区域优化约束扩散轨迹的KL散度，实现大规模并行在线强化学习中的稳定训练，在73个任务中优于或持平基线。

链接：https://arxiv.org/abs/2606.15260

作者：Huy Le, Onur Celik, Denis Blessing, Tai Hoang, Claas A Voelcker, Axel Brunnbauer, Felix Richter, Michael Volpp, Gerhard Neumann

英文摘要： Reinforcement learning with massively parallel simulations has become a standard framework for developing robust, deployable policies; however, most existing approaches still rely on simple Gaussian policy parameterizations. Diffusion models provide a more expressive policy class and have shown strong performance on challenging control problems, yet most diffusion-based RL methods are designed for offline or off-policy training. In this work, we ask whether diffusion policies can be trained effectively in the massively parallel, on-policy regime. To this end, we introduce Trust-region Diffusion Policies (TruDi), which enables diffusion policies for on-policy RL with massively parallel simulations. This setting is particularly challenging because the data distribution changes quickly across updates, making stable training with complex policies difficult. TruDi addresses this by integrating a trust-region optimization rule to enforce a KL-divergence constraint over the entire diffusion trajectory. Empirically, we evaluate TruDi on a diverse set of 4 massively parallel RL benchmarks comprising a total of 73 tasks. Across these tasks, TruDi consistently outperforms or is on-par with strong baselines on standard tasks and achieves clear gains on more challenging humanoid control tasks, establishing a strong new baseline for massively parallel on-policy RL.

AI 总结：利用深度强化学习和AlphaZero风格自我对弈，在LLL原始动作空间中学习更优的格基约简策略，训练于8维格但可零样本泛化至32维。

链接：https://arxiv.org/abs/2606.15301

机构：FAIR, Meta Superintelligence Labs（Meta超级智能实验室FAIR）； Sorbonne Université CNRS, LIP6（索邦大学CNRS/LIP6）； EPITA, EPITA Research Lab (LRE)（EPITA研究实验室(LRE)）

作者：Mohamed Malhou, Kristin Lauter, Ludovic Perret

英文摘要：The Lenstra-Lenstra-Lovász (LLL) algorithm is a seminal contribution to computer science used for lattice basis reduction, yet its polynomial-time outputs produce bases that are far from optimal as the dimension grows. We show that deep reinforcement learning can discover strictly superior, generalizable reduction strategies by interacting with the primitive action space of LLL. We formulate lattice reduction as a single-player Markov Decision Process (MDP) and train a deep residual network using an AlphaZero-style self-play pipeline augmented with adaptive-horizon MCTS (Monte Carlo Tree Search), which couples multi-step network predictions with an entropy-gated expansion mechanism. The resulting policy, DeltaStar, is trained exclusively on small $8$-dimensional $q$-ary lattices and requires fewer primitive row operations than LLL. Crucially, it generalizes zero-shot to unseen moduli and higher dimensions up to $n=32$ without retraining.

AI 总结：使用GRPO微调LLM，结合Wikipedia修订工具获取实时信息，预测未来事件，使1.5B参数模型性能超越Claude Sonnet 3.5。

链接：https://arxiv.org/abs/2606.15917

作者：Amit Arnold Levy

英文摘要：We use Group Relative Policy Optimization (GRPO), a recently devised sample and memory efficient reinforcement learning method, to finetune pretrained LLMs in the range of 1.5B to 14B parameters equipped with the ability to get current information through the use of a Wikipedia revisions tool, or news summaries, to forecast real events beyond the knowledge cutoff of the LLM, as well as problems made to simulate different aspects of the dynamics of that training. We use the results of these experiments to comment on the scaling capability of LLMs for forecasting, as well as classify how judgmental forecasting fits into the verifiable/unverifiable domain taxonomy, considering the impact of the inherent aleatoric uncertainty when forecasting future events (e.g. the roll of a die). As a result of the GRPO training, we manage to bring a 1.5B parameter transformer (Qwen 2.5 1.5B) to forecasting performance superior to Claude Sonnet 3.5 over the same dataset as measured by cross entropy from the market agreed probabilities. We also discuss various dead ends on the path to this result.

AI 总结：针对非均匀更新频率下的蒙特卡洛乐观策略迭代，本文通过一个三状态MDP反例证明标量步长非均匀异步值迭代可能不收敛，并揭示了各向异性畸变导致的切换吸引环。

链接：https://arxiv.org/abs/2606.15978

作者：Yuanlong Chen

英文摘要：Tsitsiklis proved convergence of Monte Carlo optimistic policy iteration under a uniform update structure and identified nonuniform update frequencies as a delicate obstruction. We give a certified negative answer for the natural scalar-stepsize, unnormalized asynchronous state-value recursion with fixed nonuniform state-selection probabilities. In a three-state, two-action discounted MDP, the nonuniform update frequencies induce a diagonally scaled greedy-policy mean field with a certified nonconstant attracting hybrid periodic orbit. With a bounded unbiased geometric-horizon estimator and Robbins--Monro stepsizes, the original stochastic recursion remains trapped near the cycle with positive probability and therefore fails to converge. The example pinpoints a geometric obstruction: uniform sampling gives radial residual contraction, whereas scalar nonuniform sampling anisotropically distorts the residual dynamics and can generate switched attracting cycles.

AI 总结：通过令牌级梯度动力学分析GRPO的不稳定性，提出仅更新正优势完成的WAPO算法，在数学推理和多跳QA任务中提升训练稳定性并匹配或超越基线。

链接：https://arxiv.org/abs/2606.16154

作者：Prasanth YSS, Zhichen Ren, Rasa Hosseinzadeh, Ilan Gofman, Yuqi Chen, Zhaoyan Liu, Guangwei Yu, Jesse C. Cresswell, Satya Krishna Gorti

英文摘要： Reinforcement learning with verifiable rewards (RLVR) improves language-model reasoning, but GRPO-style optimization remains prone to collapse. We analyse this instability through token-level gradient dynamics, deriving a taxonomy that predicts how updates affect next-token probabilities and entropy. The taxonomy shows that stability depends jointly on the advantage sign and token distribution under the current policy. Motivated by this finding, we propose Winner Advantage Policy Optimization (WAPO), a simple online clipped policy-gradient objective that updates only on positive-advantage completions. Across mathematical reasoning and multi-hop QA benchmarks, WAPO improves training stability and matches or outperforms baselines across multiple model families. Full code can be found at this https URL.

AI 总结：提出GERS方法，通过双层优化利用标量验证反馈调整奖励函数，在限制轨迹访问下提升强化学习在未见环境中的泛化性能。

链接：https://arxiv.org/abs/2606.16236

机构：University of Tsukuba（筑波大学）； RIKEN Center for Advanced Intelligence Project（理化学研究所革新智能综合研究中心）

作者：Ekasit Usaratniwart, Xilin Gao, Marc Ong, Youhei Akimoto

英文摘要：Reinforcement learning (RL) often suffers from performance degradation when deployed in environments that differ from those encountered during training. Existing techniques such as domain randomization (DR) mitigate this, but require access to diverse training environments and full trajectory observability, assumptions that fail in privacy-preserving or restricted scenarios where only scalar performance metrics are available. We propose Generalization via Evolutionary Reward Shaping (GERS), a bilevel optimization approach to improve generalization on unseen test environments using only scalar feedback from validation environments. At the lower level, an RL agent guided via a reward function shaped by the upper level learns a policy on a limited set of training environments with accessible trajectory data; at the upper level, CMA-ES optimizes the reward shaping parameters to maximize the cumulative unshaped reward on separate validation environments for which trajectory access is unavailable. Results on continuous control tasks indicate that GERS outperforms the standard RL baseline on unseen test environments. GERS performance is comparable to DR, despite DR treating the combined set of training and validation environments of GERS as a single training set that requires trajectory access, whereas GERS cannot access validation trajectories. These results confirm that GERS effectively enhances generalization under restricted data access constraints.

AI 总结：提出FlowMPC框架，结合流匹配模仿策略与学习的世界模型，通过MPPI规划提升测试时性能，在ManiSkill操作任务中显著提高成功率。

链接：https://arxiv.org/abs/2606.16286

机构：Stanford University（斯坦福大学）

作者：Chandon Hamel

英文摘要：Flow Matching (FM) is a powerful approach for behavior cloning in multimodal action spaces [Jiang et al., 2025], but because it is not trained to directly maximize expected return, there is still room to improve how FM policies act at test time. This work investigates whether a learned world model can improve FM policies by enabling Model Predictive Path Integral (MPPI) planning over candidate action sequences proposed by the policy. Building on TD-MPC2 [Hansen et al., 2024], I introduce FlowMPC, a framework that combines an imitation-learned FM policy with a learned world model for test-time planning in ManiSkill manipulation tasks [Tao et al., 2025]. Across PickCube and PickSingleYCB, adding the world model improved performance over the FM policy alone, with especially clear gains in end-of-episode success. These results suggest that world-model-based planning can effectively complement flow-based imitation policies without modifying the FM training objective.

AI 总结：提出扩散软演员-评论家方法，结合保守Q学习与扩散模型，在离线强化学习中优化无人机轨迹与调度，降低能耗并提升公平性，性能优于现有算法。

链接：https://arxiv.org/abs/2606.16331

机构：Centre for Wireless Communications (CWC), University of Oulu（奥卢大学无线通信中心（CWC））

作者：Eslam Eldeeb, Hirley Alves

英文摘要：The integration of generative artificial intelligence with wireless communication and signal processing systems has opened new avenues for intelligent, data-driven decision-making in future 6G networks. This work proposes a diffusion soft actor-critic (Diffusion-SAC) approach that leverages offline reinforcement learning (RL) enhanced by denoising diffusion probabilistic models (DDPMs) to optimize trajectory and scheduling control in unmanned aerial vehicle (UAV) networks. While offline RL methods, such as conservative Q-learning (CQL), can learn from static datasets, they often struggle to generalize in low-data or dynamic conditions. To address this, we combine the robustness of CQL with the generative power of diffusion models, enabling expressive and signal-aware policy learning that generalizes beyond behavior policies. Applied to a UAV-assisted wireless network, the proposed framework minimizes transmission energy and improves fairness among devices. Simulations show that Diffusion-SAC outperforms standard offline RL baselines, achieving more stable convergence and higher rewards even with limited datasets. The method enhances data efficiency, reduces energy consumption, and increases throughput by more than 35 % compared to existing algorithms, demonstrating its potential for robust policy learning in next-generation wireless control systems.

AI 总结：提出BRICKS-WM框架，将全局动力学分解为通过潜在接口交互的独立模块（如智能体和背景），实现冻结背景模块跨智能体重用，避免从头训练。

链接：https://arxiv.org/abs/2606.16489

机构：National Key Laboratory for Novel Software Technology, Nanjing University, China（南京大学计算机软件新技术国家重点实验室）； School of Artificial Intelligence, Nanjing University, China（南京大学人工智能学院）； School of Intelligence Science and Technology, Nanjing University, China（南京大学智能科学与技术学院）

作者：Shaowei Zhang, Jiahan Cao, Xunlan Zhou, Shenghua Wan, De-Chuan Zhan

英文摘要：Model-based Reinforcement Learning (MBRL) has achieved remarkable success in continuous control by leveraging latent world models. However, prevailing approaches typically rely on monolithic latent dynamics, entangling environment dynamics into a coupled process. This coupling severely limits reusability: altering the agent necessitates retraining the entire world from scratch, even if the environment remains constant. To address this, we introduce BRICKS-WM (Building Reusability via Interface Composition Kinetics for Structured World Models), a framework for the modular assembly of structured world models. Driven by the insight that the physical world is composed of independent entities, we posit that global dynamics can be modeled as a composition of distinct dynamical modules interacting via latent interfaces. As a minimal instantiation, we factorize the latent state space into an actuated Agent module and an external Background module, bridged by a learned latent interface. Unlike prior object-centric methods that prioritize visual segmentation, BRICKS-WM enforces a functional separation in transition dynamics, ensuring that background dynamics remains agnostic to the agent's dynamics. Empirically, BRICKS-WM achieves control performance comparable to strong monolithic baselines when trained from scratch, and enables the reuse of frozen background dynamics across agents.

AI 总结：提出daVinci-kernel框架，通过强化学习联合训练技能选择、策略生成和技能总结三个智能体，共享LLM骨干，实现GPU内核优化，在KernelBench上超越先前最优模型。

链接：https://arxiv.org/abs/2606.16497

作者：Dayuan Fu, Mohan Jiang, Tongyu Wang, Dian Yang, Jiarui Hu, Liming Liu, Jinlong Hou, Pengfei Li

英文摘要：GPU kernel optimization represents a paradigm where functional correctness is assumed and execution efficiency is the objective. We present daVinci-kernel, a reinforcement learning framework that couples skill discovery with skill exploitation through a dynamically evolving skill library. daVinci-kernel jointly trains three agents sharing one LLM backbone: a Skill Selection Agent that retrieves relevant techniques via BM25 and LLM reranking, a Policy Agent that generates multi-turn CUDA/Triton kernels conditioned on selected skills, and a Skill Summary Agent that distills successful rollouts into reusable skills. Candidate skills are added only after execution-based verification confirms reproducible speedups. All three agents share a single LLM backbone, are initialized via a structured SFT cold start on diversity-filtered data, and are then jointly optimized end-to-end with multi-turn REINFORCE and per-agent advantage estimation. On KernelBench, daVinci-kernel-14B achieves 37.2%, 70.6%, and 32.2% on Level 1, Level 2, and Level 3 under the Fast$_1$ threshold, outperforming the strongest prior RL-trained model, this http URL -14B.

AI 总结：提出方向条件策略（DCP），通过共享InfoNCE表示将目标达成分解为子目标评分和方向条件动作，理论证明方向充分性、训练与部署一致性及可控子空间失效条件，在九个环境中优于对比RL。

链接：https://arxiv.org/abs/2606.16515

作者：Swaminathan S K, Damiya Gondha, Theyanesh Eswaramoorthy Rajahkrishnan, Aritra Hazra

英文摘要：Hamilton-Jacobi-Bellman theory implies that the optimal goal-conditioned action depends on the goal only through the gradient of the goal-reaching distance at the current state, yet standard online GCRL still conditions the actor on the raw goal -- a signal that is geometrically uninformative when the goal is far from the data distribution. We propose Direction-Conditioned Policies (DCP), a fully online method that decomposes goal-reaching into two components sharing one InfoNCE representation $\psi$: a subgoal-scoring step that selects a visited state $z_t$ aligned with the final goal $g$ in $\psi_g$, and a direction-conditioned actor that consumes the unit direction $d_t$ and magnitude $r_t$ from $\psi(s_t)$ to $\psi(z_t)$. The two components train jointly, factor cleanly at deployment (subgoal scoring is removed, while direction conditioning remains with $g$ in place of $z_t$), and admit independent modification at the same $(d_t,r_t)$ interface. We prove three results. First, direction sufficiency under HJB: the optimal action under control-affine dynamics depends on the goal only through the value gradient. Second, a quantitative bound showing that, under mild conditions on the learned representation and assuming the scoring rule returns an on-path $z_t$, the actor's conditioning input at training and at deployment coincide up to representation error and geodesic slack. Third, a controllable-subspace characterization of when directional conditioning fails. Across nine environments, DCP improves over Contrastive RL on most final metrics, with the largest gains on manipulation and obstacle-interaction tasks; a qualitative analysis of the learned $\psi$-distance landscape shows the contrastive representation behaves as an online quasimetric encoding environment topology, and the single failure case (AntSoccer) localizes to a learned-gradient pathology that the theory anticipates.

AI 总结：受婴儿自发运动噪声启发，提出一种在RL训练中逐步增加时间自相关的探索噪声机制，实验表明其能产生结构化探索行为并提高学习效率。

链接：https://arxiv.org/abs/2606.16590

机构：Frankfurt Institute for Advanced Studies（法兰克福高等研究所）； School of Computer Science and Engineering, University of New South Wales（新南威尔士大学计算机科学与工程学院）； Escuela de Ingeniería, Universidad Central de Chile（智利中央大学工程学院）； Faculty of Electrical Engineering, Czech Technical University（捷克理工大学电气工程学院）

作者：Francisco M. López, Markus R. Ernst, Francisco Cruz, Matej Hoffmann, and Jochen Triesch

英文摘要： Exploration in deep reinforcement learning (RL) is commonly implemented as temporally uncorrelated white noise. However, recent works show that temporally correlated colored noise can improve exploration efficiency by producing smooth trajectories with better coverage of the state space. We inquire whether action noise inspired by infant spontaneous movements can also improve exploration in deep RL. We find that the power spectral densities of babies' end-effector velocities follow a colored noise process where the spectral exponent increases with age. Inspired by this developmental pattern, we introduce a mechanism that progressively increases the temporal auto-correlation of exploration noise during RL training, matching the infant statistics. Experiments across several RL environments show that infant-inspired noise produces structured exploratory behavior and can improve learning efficiency compared to conventional exploration strategies. These findings suggest that human motor and cognitive development can provide useful guidance for designing learning mechanisms in artificial agents. Our code is available at this https URL.

AI 总结：研究多种延迟模型下的随机线性赌博机，给出近最优遗憾界，揭示延迟与线性结构交互的维度影响。

链接：https://arxiv.org/abs/2606.16656

作者：Ofir Schlisselberg, Mengxiao Zhang, Yishay Mansour

英文摘要：We study stochastic linear bandits with delayed feedback under several delay models and establish near-optimal regret guarantees. Our results identify when delayed linear bandits exhibit the same qualitative behavior as multi-armed bandits (MAB), and when the linear structure creates fundamentally new challenges. Specifically, (1) for \emph{loss-independent delays}, where the delay does not depend on the realized loss (but potentially depends on the arm), we show that delays incur only an additive regret penalty. Under stochastic delays, this penalty scales with the expected delay, while under adversarial delays, it scales with the maximum number of outstanding observations. Notably, both delay penalties are dimension-free, improving upon the state-of-the-art results; (2) for \emph{loss-dependent delays}, we show that linear bandits are substantially harder than MAB: unlike in MAB, we prove matching (up to log factors) upper and lower bounds in linear bandits, whose delay penalty depends on the square root of the dimension. (3) for the \emph{delay-as-payoff model}, a special case of loss-dependent delay, we show that the optimal MAB guarantee, which depends only on the delay of the optimal arm, is also unattainable in linear bandits. Together, these results provide a sharp characterization of how delayed feedback interacts with linear generalization.

AI 总结：针对弱通信平均奖励MDP，首次从单条轨迹建立有限样本复杂度保证，提出无模型方法，值函数和策略方法分别达到$\widetilde{O}(1/\varepsilon^2)$和$\widetilde{O}(1/\varepsilon^4)$的样本复杂度。

链接：https://arxiv.org/abs/2606.16729

作者：Jongmin Lee, Ernest K. Ryu, Vaneet Aggarwal

英文摘要：While there is an extensive body of work characterizing the sample complexity of discounted cumulative-reward MDPs, finite sample analyses for average-reward MDPs have been limited, and most existing works rely on restrictive assumptions such as ergodicity or access to a generative model. In this work, we establish the first finite sample complexity guarantees from a single trajectory for weakly communicating average-reward MDPs. To this end, we study the dynamics of a single trajectory in weakly communicating MDPs and based on this analysis, we develop novel model-free methods. Notably, our value-based and policy-based methods provide finite sample complexity guarantees of $\widetilde{O}(1/\varepsilon^2)$ and $\widetilde{O}(1/\varepsilon^4)$ from a single trajectory in weakly communicating MDPs, respectively. Furthermore, we introduce the first model-free method that requires no prior knowledge of problem-dependent quantities for communicating MDPs.

AI 总结：针对平均奖励准则下的离散时间无限时域均值场博弈，提出基于最大因果熵的逆强化学习方法，通过占据测度框架统一处理有限维线性奖励和无限维RKHS奖励，并设计梯度上升算法实现策略恢复。

链接：https://arxiv.org/abs/2606.16759

机构：Bilkent University（比尔肯大学）； Özyeğin University（厄齐金大学）

作者：Şevket Kaan Alkır, Naci Saldı, Berkay Anahtarcı, Can Deha Karıksız

英文摘要：We study inverse reinforcement learning for discrete-time, infinite-horizon mean-field games (MFGs) under an average-reward criterion. Expert demonstrations are assumed to arise from a stationary mean-field equilibrium under an unknown reward, and the goal is to recover a policy explaining the observed behaviour via the maximum causal entropy principle. We formulate the inverse problem by enforcing consistency with the expert mean-field term and long-run feature expectations, treating two reward classes within a unified occupation-measure framework. For finite-dimensional linear rewards, we give a convex dual reformulation with an explicit log-partition objective, and prove smoothness and curvature properties justifying constant-step-size gradient descent. For infinite-dimensional RKHS rewards, we develop a Lagrangian relaxation whose inner-maximising policy is characterised by a soft Bellman equation. The main obstacle is the absence of a discount-factor contraction. We resolve this by introducing a minorisation-based sub-stochastic kernel that yields a strict contraction of the soft Bellman operator. We establish Fréchet differentiability and Lipschitz smoothness of the log-likelihood score, leading to a gradient ascent algorithm with convergence guarantees. Two numerical examples, a malware-spread MFG and an RKHS-based consumer-choice model, show that the recovered policies closely match expert behaviour.

AI 总结：提出GD$^2$PO算法，通过冲突感知过滤机制屏蔽奖励不一致的rollout，并结合查询级重加权，解决多奖励优化中的信号抵消问题，提升RL训练效率。

链接：https://arxiv.org/abs/2606.16771

机构：Qwen Large Model Application Team, Alibaba（阿里巴巴通义千问大模型应用团队）； Renmin University of China（中国人民大学）； Peking University（北京大学）； ETH Zürich（苏黎世联邦理工学院）； University of Zurich（苏黎世大学）； The Chinese University of Hong Kong（香港中文大学）

作者：Haotian Liu, Yihao Liu, Jingwei Ni, Siyuan Huang, Xinpeng Liu, Pengyu Cheng, Jiajun Song, Ruijin Ding, Junfeng Li, Zhechao Yu, Mengyu Zhou, Hongteng Xu, Xiaoxi Jiang, Guanjun Jiang

英文摘要：As LLMs advance, post-training reinforcement learning (RL) increasingly relies on multi-dimensional rewards to cultivate comprehensive capabilities. This shift demands new algorithms capable of optimizing diverse and potentially competing objectives simultaneously. To address this, existing methods such as Group reward-Decoupled Policy Optimization (GDPO) decompose the overall score into independent reward groups, then compute the RL loss separately within each group. However, this strategy still encounters multi-reward conflicts: a single rollout can yield positive advantages on certain reward dimensions but negative ones on others, causing opposing signals to cancel each other out during aggregation, further hindering RL training efficiency. Inspired by Dynamic sAmpling Policy Optimization (DAPO), which improves RL training efficiency by filtering out ineffective rollouts with near-zero advantages, we propose Group-Dynamic reward-Decoupled Policy Optimization (GD$^2$PO). Specifically, GD$^2$PO employs a conflict-aware filtering mechanism to mask out rollouts suffering from severe reward-wise disagreement. By preventing conflicting signals from canceling each other out, this masking strategy preserves and enhances the magnitude of effective RL advantages, thereby significantly accelerating learning efficiency. Furthermore, we introduce query-level reweighting to dynamically adjust the update intensity of each query based on its overall reward consensus. Experiments on various multi-reward scenarios, including tool calling and human preference alignment, demonstrate that GD$^2$PO consistently and significantly outperforms existing baselines. The code is available at this https URL.

AI 总结：研究连续时间随机控制中Q学习的算子核心，通过分析扩散设置下Bellman最优性目标的正则性和逼近复杂度，提出适应混合正则性的张量积DeepONet架构，并给出显式逼近和资源界限。

链接：https://arxiv.org/abs/2606.16846

机构：Peking University（北京大学）

作者：Qian Qi

英文摘要：We study the operator-theoretic core of Q-learning in continuous-time stochastic control with continuous states and actions. In value-based reinforcement learning, each Q-learning or DQN update is built from a Bellman optimality target; our analysis isolates this target in a diffusion setting and studies its regularity and approximation complexity. Under uniform ellipticity and Hölder-regular coefficients, we show that a Bellman update maps bounded inputs into an anisotropic regularity class, smoothing the state variable while leaving only Lipschitz dependence on the action variable. This yields a compact family of Bellman iterates and motivates a tensor-product DeepONet architecture adapted to the mixed regularity of the problem. We then derive explicit approximation and resource bounds, together with a stiffness--complexity trade-off as the time step $\delta \to 0$. The resulting theory makes a direct contribution to Q-learning theory at the level of Bellman target regularity and approximation in continuous stochastic control. At the same time, we do not claim a full convergence theorem for practical sampled Q-learning with exploration, replay, and stochastic gradient updates.

AI 总结：提出一种统一因果起源分类法，将强化学习中的分布偏移按因果来源（内部/外部）和时间边界（显式/隐式/混合）分类，统一了分布内/外泛化与非平稳性分析。

链接：https://arxiv.org/abs/2606.16933

机构：IMT Atlantique（IMT大西洋）； Flinders University（弗林德斯大学）； IRL Crossing； Priori Analytica； CNRS（法国国家科学研究中心）

作者：Ardianto Wibowo, Paulo E Santos, Amer Baghdadi, Matthew Stephenson, Karl Sammut, Jean-Philippe Diguet

英文摘要：Reinforcement learning (RL) systems often degrade when operating conditions differ from those previously encountered, reflecting distributional shifts in the underlying data-generating process. Such shifts may occur between training and evaluation, as in In-Distribution (ID) and Out-of-Distribution (OOD) generalization, or within non-stationary settings where environment dynamics evolve over time. However, the formal relationship between these views remains unclear, and existing work mainly focuses on mitigation rather than the causal origin of shift within the agent-environment interaction. This work develops a unified causal-origin taxonomy that characterizes sources of distributional shift in RL and relates ID/OOD generalization to non-stationary settings. We transfer the classical dataset-shift principle from supervised learning to RL by reformulating distributional shift in terms of the generative interaction process. Using a Partially Observable Markov Decision Process (POMDP), we decompose the interaction into structural components, including the state distribution, observation process, policy, reward, and transition dynamics, together with the shifted-time boundary. The proposed taxonomy distinguishes internal, agent-driven, and external, environment-driven, distributional shifts. The shifted-time boundary perspective further characterizes explicit, implicit, and hybrid shifts. This formulation unifies ID/OOD generalization and non-stationarity as structured changes in the underlying process. We also introduce an evaluation framework for measuring shift impact and adaptation through performance degradation and recovery metrics. By grounding distributional shift in the causal-origin structure of RL, this work supports systematic analysis of robustness under distributional shift.

AI 总结：提出ExpRL方法，利用人类编写的问答数据作为奖励支架，通过密集奖励强化推理过程中的部分进展和有用行为，在数学推理任务上优于SFT、稀疏奖励GRPO和自蒸馏，并为后续稀疏奖励RL提供更好的初始化。

链接：https://arxiv.org/abs/2606.17024

机构：Stanford University（斯坦福大学）； Carnegie Mellon University（卡内基梅隆大学）； OpenAI； Rogo

作者：Violet Xiang, Amrith Setlur, Chase Blagden, Nick Haber, Aviral Kumar

英文摘要：Sparse reward reinforcement learning (RL) has become a standard tool for improving LLM reasoning, but its success depends critically on the coverage present in the base model. In practice, models are often primed for RL through \emph{mid-training} on curated reasoning traces that teach useful primitive skills such as decomposition, verification, or self-correction. Although effective, this strategy requires manually specifying what the model should learn, and it remains unclear whether such primitive coverage is enough for much harder problems, which require combining these skills into broader solution strategies. We study a more automated approach: \emph{RL-based mid-training} using large corpora of human-written question-answer data. Rather than treating reference solutions as targets to imitate, our method, ExpRL, uses them as \emph{reward scaffolds}: references are hidden from the policy and used only to construct problem-specific grading rubrics for judging on-policy reasoning traces. The policy samples from the original problem prompt, while an LLM judge compares the sampled reasoning trace against the reference solution and assigns outcome-level or process-level dense rewards. This lets ExpRL reinforce partial progress, useful intermediate reductions, and productive reasoning behaviors that sparse final-answer rewards often fail to upweight. On challenging math reasoning tasks, ExpRL yields stronger RL priming than SFT, sparse-reward GRPO, and self-distillation, and provides a better initialization for subsequent sparse-reward RL. Additional mixed-domain experiments further suggest that ExpRL can extend beyond the original math-only setting.

AI 总结：提出时间差分（TD）目标函数，通过将扩散过程视为马尔可夫奖励过程并利用强化学习中的策略评估，强制去噪轨迹上的跨时间一致性，显著提升少步采样下的生成质量。

链接：https://arxiv.org/abs/2606.15048

作者：Qizhen Ying, Yangchen Pan, Victor Adrian Prisacariu, Junfeng Wen

英文摘要：Diffusion models are typically trained with objectives that focus on local denoising targets at individual time steps (or adjacent pairs), which do not enforce consistency between predictions along the denoising trajectory. This lack of cross-time consistency can degrade performance, especially for few-step samplers. We introduce a temporal difference (TD) objective that penalizes inconsistency of the model's multi-step progress along the denoising path. By reformulating the diffusion process as a Markov reward process and casting denoising as a policy evaluation problem in reinforcement learning, we derive a unified TD approach that applies to both discrete- and continuous-time diffusion formulations. We further propose a principled sample-based reweighting method that stabilizes training. Empirically, we show that using our TD training can significantly improve sample quality measured by FID, with stronger advantages when the number of sampling steps is small, highlighting its practical utility under low-computation-budget scenarios. We provide ablation studies to justify our design choices, including pairwise loss reweighting, regularization weight, and one-step stride. Overall, our TD approach can be a general drop-in that enforces cross-time consistency and improves generation quality across different diffusion generative models.

AI 总结：提出TimeMoDE框架，结合扩散Transformer与专家混合，通过领域提示和扩散时间步信号，在数据稀缺场景下生成高质量时间序列，显著优于现有方法。

链接：https://arxiv.org/abs/2606.15172

作者：Zihao Yao, Qi Zheng, Jiankai Zuo, Yaying Zhang

英文摘要：Synthesizing realistic time series with generative models has wide-ranging applications in real-world scenarios. Despite recent progress, most existing methods are trained under the assumption of abundant training data, which substantially limits their effectiveness in data-scarce settings. In this paper, we propose TimeMoDE, a novel framework that integrates Diffusion Transformers with Mixture-of-Experts to exploit both domain adaptability and diffusion-stage awareness for time series generation under data scarcity. It is pre-trained on a large-scale collection of multi-domain datasets to extract domain-agnostic temporal representations and domain-specific information benefiting generalization during fine-tuning. We propose Domain Prompts to condition expert assignment for indistinguishable noised tokens, mitigating the limitations of capturing inter-dataset relationships. Moreover, we incorporate diffusion timestep signals to equip the experts with awareness of time series degradation variations, facilitating adaptive calibrate to stage-dependent denoising requirements. Extensive experiments demonstrate that TimeMoDE outperforms existing methods under diverse low-data settings. It establishes an innovative paradigm for advanced time series few-shot generation.

AI 总结：本文通过分析泛化误差的三个关键因素，提出SemDLM+模型，通过全局转移和语义频率惩罚解决语义盆地问题，在LM1B和OpenWebText上提升了训练动态和生成质量。

链接：https://arxiv.org/abs/2606.15327

机构：Alibaba Group（阿里巴巴集团）； Fudan University（复旦大学）； University College London（伦敦大学学院）； Nanyang Technological University（南洋理工大学）； University of Oxford（牛津大学）

作者：Keyue Jiang, Yuxiang Wang, Yanan Zhao, Xiang Yu, Qifang Zhao, Bohan Tang, Baojian Zhou, Yanghua Xiao, Lin Qu, Xiaoxiao Xu

英文摘要： Diffusion Language Models (DLMs) have demonstrated strong scaling capacity as alternatives to autoregressive language models. However, their performance is highly sensitive to the choice of transition kernels, and poorly designed kernels can lead to issues like training instability, slow convergence, and biased sampling. In this paper, we study this sensitivity through a principled analysis of generalization error and identify three critical factors: asymptotic bias (difficulty in approximating the posterior distribution), exposure bias (error propagation during sampling), and optimization variance induced by kernel dispersion. We further compare different transition kernels: masking diffusion yields sparse and easier posterior-approximation targets, while uniform diffusion provides stronger sampling-side repair but induces harder approximation. Motivated by this trade-off, we revisit a previously overlooked variant, semantic DLM (SemDLM), where the transition kernel corrupts tokens to neighborhoods that are semantically similar. Our theory suggests that SemDLM can serve as a plausible middle ground by reducing the posterior approximation difficulty of uniform diffusion while retaining repair ability. However, we find that SemDLM suffers from a semantic basin problem, where sampling repeatedly stays within a semantic region and produces low-diversity text. To address this, we propose SemDLM+, which adds a global transition and a semantic-frequency penalty during sampling. Experiments on LM1B and OpenWebText show that SemDLM+ improves training dynamics and achieves competitive language modeling and generation quality with satisfactory diversity.

AI 总结：针对截断签名反演的病态问题，提出概率框架，采用签名条件流匹配模型学习路径的条件分布，并推导线性统计下的贝叶斯最优误差作为理论基线。

链接：https://arxiv.org/abs/2606.15332

机构：ECE & IPAI, Seoul National University（首尔大学电气与计算机工程系 & 人工智能研究所）； Department of Statistics, Purdue University（普渡大学统计系）

作者：Junoh Kang, Kiseop Lee, Bohyung Han

英文摘要：The signature transform is a principled feature map for continuous-time paths, valued for its uniqueness and universality. Recovering a path from its truncated signature is, however, structurally ill-posed because the truncated signature map is not injective. We therefore reframe truncated signature inversion as a probabilistic problem -- learning the conditional distribution of a path given its truncated signature -- and adopt a signature-conditioned flow matching model as a practical estimator. This probabilistic formulation elucidates the fundamental difficulty of inversion: Bayes reconstruction error quantifies the irreducible uncertainty remaining after conditioning on a statistic. We derive the Bayes-optimal error under linear statistics, obtaining a closed form for log-GBM and numerically tractable formulas for log-fBM and OU, yielding a concrete theoretical baseline for model validation. This baseline upper-bounds the Bayes error under truncated-signature conditioning, since truncated signatures provide richer information than linear statistics. Experiments show that empirical reconstruction errors under linear-statistics conditioning faithfully align with the theory-derived baseline, while errors decrease when the statistic is replaced with truncated signatures. Moreover, generated paths faithfully recover the conditioning signature while preserving key distributional and temporal structures, indicating that the estimator is well-calibrated to the target conditional distribution. Together, these results establish a well-posed probabilistic framework for truncated-signature inversion, with applicability demonstrated on real financial data beyond the parametric process families covered by theory.

AI 总结：提出DiRecT算法，通过随机最优控制仅在最终干净轨迹上施加约束，避免中间去噪步骤过度约束，实现安全扩散规划，提升安全性和任务性能。

链接：https://arxiv.org/abs/2606.15359

机构：MIT（麻省理工学院）

作者：Paolo Giaretta, Zeyang Li, Navid Azizan

英文摘要：Diffusion models have emerged as powerful tools for planning and control by learning multimodal distributions over actions and trajectories. Yet reliable inference-time safety enforcement remains a key barrier to their deployment in safety-critical tasks. Existing approaches typically project each denoising iterate onto the feasible set, even though constraints are defined only on the final clean trajectory. Enforcing feasibility on noisy intermediate samples can therefore overconstrain the sampling dynamics, substantially degrading sample quality. To address this limitation, we introduce DiRecT (Diffusion-based planning via Receding-horizon denoising with Terminal constraints), a training-free algorithm for constrained sampling from diffusion models via stochastic optimal control (SOC). DiRecT enforces constraints only on the final clean sample, avoiding unnecessary restrictions on the intermediate denoising dynamics. Inspired by model predictive control, we derive a principled receding-horizon surrogate for the otherwise intractable constrained SOC formulation, yielding an efficient algorithm that cleanly separates stochastic denoising from constraint satisfaction, progressively steering samples toward feasible final trajectories without distorting the learned diffusion dynamics. Furthermore, DiRecT is highly flexible: it can leverage off-the-shelf or domain-specific optimizers, incorporate priors over environment dynamics, and optimize additional soft rewards. Extensive experiments on safe planning benchmarks demonstrate that DiRecT substantially improves deployment safety and task performance over existing diffusion-based planning baselines.

AI 总结：提出PHINN框架，利用动态Betti曲线作为条件信号和持久景观损失保持同调一致性，在金融、流行病和多模态基准上拓扑保真度优于统计和扩散基线。

链接：https://arxiv.org/abs/2606.15452

机构：Defense.Codes (a DBA of CapaCloud Corp)（Defense.Codes（CapaCloud Corp 的商用名））

作者：Emre Yusuf, Ren Takahashi, Jayabrata Bhaduri

英文摘要： Rare events in time series are critical to model but hard to learn due to data scarcity. Current generative models struggle with extreme values. We observe that rare events leave distinct topological fingerprints - transitions in Betti numbers from point-cloud embeddings - that are more stable and discriminative than statistical moments. We introduce PHINN, a flow-matching framework using dynamic Betti curves as conditioning signals and a persistence landscape loss for homology consistency. It scales to multivariate data, includes a natural-language interface to set Betti targets, supports cross-domain meta-learning and few-shot generation, and provides certified adversarial robustness. On financial, epidemiological, and multi-modal benchmarks, PHINN outperforms statistical and diffusion baselines in topological fidelity (beta-RMSE down 41-63%, transition accuracy up 84%) and matches jump-diffusion models in tail coverage while exceeding them in shape fidelity. All results have 95% confidence intervals.

AI 总结：本文在生成流网络框架下，推导了策略梯度算法并首次应用近端策略优化，提升了离散概率分布采样的收敛速度和数据效率。

链接：https://arxiv.org/abs/2606.15793

机构：HSE University（高等经济学院）； Constructor University（康斯特大学）； CMAP, CNRS, École polytechnique, IPP（CMAP，CNRS，巴黎综合理工学院，IPP）

作者：Anna Zykova-Myzina, Timofei Gritsaev, Daniil Tiapkin, Nikita Morozov

英文摘要：This paper explores policy gradient algorithms for training stochastic policies to sample from structured discrete probability distributions under the Generative Flow Network (GFlowNet) framework. Building on extensive theoretical connections between GFlowNets and entropy-regularized reinforcement learning, we derive equivalents of standard policy gradient algorithms for training GFlowNets, as well as experimentally explore their various methodological aspects, including baseline training and advantage estimation. Most importantly, our work is the first to derive and successfully apply proximal policy optimization to GFlowNets, showing its improved convergence speed and data efficiency compared to standard GFlowNet training objectives on benchmarks ranging from synthetic energies to molecular graph generation.

AI 总结：提出一种无需训练的解码框架，通过平均场变分松弛协调并行令牌更新，在单次前向传播中抑制冲突，提升质量-延迟权衡。

链接：https://arxiv.org/abs/2606.15805

机构：School of Electrical & Computer Engineering, Tel Aviv University（特拉维夫大学电气与计算机工程学院）； School of Computer Science and AI, Tel Aviv University（特拉维夫大学计算机科学与人工智能学院）； Department of Computer Science, Technion, Israel Institute of Technology（以色列理工学院计算机科学系）

作者：Tamim Zoabi, Ameen Ali, Liran Ringel, Lior Wolf

英文摘要：Discrete diffusion language models enable parallel token generation, offering a pathway to low-latency decoding. However, selecting tokens independently by marginal confidence limits effective parallelism: tokens that appear reliable in isolation can form incompatible configurations when several positions are updated at once. We introduce a training-free decoding framework that coordinates these parallel updates. At each forward pass, the method assigns a commit score to each masked position and refines these scores using pairwise interactions derived from the model's predictive distributions. A variational relaxation yields a simple fixed-point update that suppresses conflicting simultaneous commitments within a single forward pass. This mechanism allows the decoder to commit more tokens in parallel while maintaining competitive generation quality. The method is lightweight, requires no auxiliary model or retraining, and drops into existing diffusion decoding pipelines without modification. Experiments on reasoning and code-generation benchmarks show consistent improvements in the quality-latency trade-off.

AI 总结：针对去中心化扩散模型中随机专家切换的ODE采样，通过速度场分解建立Wasserstein-2距离下的收敛保证，证明N步离散化以O(N^{-1/2}+ε)速率收敛。

链接：https://arxiv.org/abs/2606.15835

机构：Peking University（北京大学）； Shanghai Jiao Tong University（上海交通大学）； MBZUAI（穆罕默德·本·扎耶德人工智能大学）； Zhejiang University（浙江大学）； Tencent（腾讯）

作者：Chencheng Tang, Xuanyu Xue, Fangyikang Wang, Chao Zhang, Hubery Yin

英文摘要：Diffusion models have achieved impressive empirical success in generative tasks, and their convergence theory is now relatively well understood. Motivated by privacy and scalability, recent decentralized diffusion architectures replace a single global velocity field with multiple local experts and a routing mechanism, yielding a sampling dynamics with stochastic expert switching that falls outside standard diffusion convergence analyses. In this work, We study a decentralized diffusion framework with stochastic velocity fields and ODE-based sampling. We establish a convergence guarantee in Wasserstein-2 distance, showing that the distribution of the $N$-step discretization converges to the analytical solution at rate $\mathcal{O}(N^{-1/2}+\varepsilon)$ in $W_2$, where $\varepsilon$ captures the neural approximation errors. To our knowledge, this is the first $W_2$ convergence result for decentralized diffusion models with an ODE-based sampling scheme.

AI 总结：提出拓扑流匹配，通过拉普拉斯漂移增强参考过程，在保留流匹配稳定性和无模拟目标的同时，捕捉底层域拓扑结构，适用于脑fMRI、洋流等结构化数据。

链接：https://arxiv.org/abs/2606.15897

作者：Kacper Wyrwal, İsmail İlkan Ceylan, Alexander Tong

英文摘要：Flow matching is a powerful generative modeling framework, valued for its simplicity and strong empirical performance. However, its standard formulation treats signals on structured spaces, such as fMRI data on brain graphs, as points in Euclidean space, overlooking the rich topological features of their domains. To address this, we introduce topological flow matching, a topology-aware generalization of flow matching. We interpret flow matching as a framework for solving a degenerate Schrödinger bridge problem and inject topological information by augmenting the reference process with a Laplacian-derived drift. This principled modification captures the structure of the underlying domain while preserving the desirable properties of flow matching: a stable, simulation-free objective and deterministic sample paths. As a result, our framework serves as a drop-in replacement for standard flow matching. We demonstrate its effectiveness on diverse structured datasets, including brain fMRIs, ocean currents, seismic events, and traffic flows.

AI 总结：提出将MCMC轨迹终止作为可学习组件，利用非循环生成流网络训练状态依赖分类器，在保证详细平衡条件下自适应停止采样，显著缩短轨迹长度并改善模式覆盖与混合。

链接：https://arxiv.org/abs/2606.16073

作者：Kirill Korolev, Nikita Morozov, Stepan Pavlenko, Esmeralda S. Whitammer, Sergey Samsonov

英文摘要：Sampling from complex, unnormalized probability densities is a fundamental challenge in Bayesian inference and probabilistic modeling. While Markov chain Monte Carlo (MCMC) methods provide asymptotic guarantees, they often suffer from slow mixing and high computational costs due to fixed or manually tuned trajectory lengths. In this work, we propose a novel framework that treats trajectory termination as a learnable component of the sampling dynamics. By framing MCMC within the theory of non-acyclic generative flow networks (GFlowNets), we train state-dependent neural classifiers to decide when a trajectory has reached a high-density region and should terminate. We theoretically establish the connection between optimal classifiers and the target density via detailed balance conditions and introduce a multilevel training scheme to facilitate exploration in complex geometries. Experimental results across various benchmark densities demonstrate that our approach significantly reduces average trajectory lengths while improving mode coverage and mixing compared to standard MCMC baselines.

AI 总结：提出MUNI框架，通过端到端多模态潜在扩散和路由训练目标，实现任意到任意生成，在条件生成上匹配或超越基线，并在无条件连贯性上取得最大优势。

链接：https://arxiv.org/abs/2606.16408

机构：KAIST（韩国科学技术院）

作者：Kyeongmin Yeo, Yunhong Min, Minhyuk Sung

英文摘要：We introduce MUNI, an end-to-end multimodal latent diffusion framework for any-to-any generation that unifies subset-conditioned cross-modal generation and unconditional joint sampling through a shared stochastic latent. Existing multimodal generative models are largely LLM-based, which limits leveraging modality-specific generators and requires text-paired data for training. Recent diffusion- and flow-based any-to-any extensions take a different direction but still rely on text-aligned embeddings, fully-paired training, or matched-dimensionality deterministic mappings. MUNI rests on two complementary contributions, one architectural and one in the training objective. First, we extend latent diffusion to multimodal any-to-any generation end-to-end: instead of the standard two-stage recipe that precomputes a frozen latent space and then fits a prior over it, MUNI jointly trains modality-specific encoders, expressive decoders, and a single shared flow-based prior under one objective. Second, we identify that the standard aggregation rules of multimodal variational inference are insufficient once coupled with a learned prior and expressive decoders. A suitable shared latent must simultaneously satisfy coherence across generated modalities, predictive sufficiency of subset latents, and minimality of the latent content. We propose a routed training objective whose structural choices align the latent with these criteria and admit a minimal-sufficiency characterization in the realizable setting. Experiments on PolyMNIST-Quadrant-Labels and a large-scale image-text-audio benchmark show MUNI matching or exceeding the strongest baselines on conditional generation while opening its largest margins on unconditional coherence. Project page: this https URL.

AI 总结：提出决策加权流匹配（DW-FM）框架，通过重加权速度回归目标对齐下游遗憾，在CVaR基准上优于标准方法。

链接：https://arxiv.org/abs/2606.16790

机构：Hong Kong University of Science and Technology（香港科技大学）； Central South University（中南大学）； Big Data Institute（大数据研究院）

作者：Jize Xie, Haomiao Wu, Qiang Chen, Xiu Su, Yi Chen

英文摘要：Conditional generative models are increasingly used as scenario generators for stochastic optimization, but standard training objectives emphasize uniform distributional fit rather than the downstream decisions induced by generated scenarios. This creates an objective mismatch: errors in statistically common regions may have little effect on decision regret, whereas errors in decision-sensitive regions can substantially change the optimal action. We propose Decision-Weighted Flow Matching (DW-FM), a regret-aligned training framework that preserves the simplicity of standard flow matching while reweighting its velocity-regression objective using decision-sensitive endpoint information. Theoretically, we connect downstream regret to pathwise velocity mismatch through a loss-induced decision discrepancy and an adjoint transport argument, yielding an ideal regret-aligned surrogate and practical endpoint-weighted objectives with regret guarantees. Empirically, we demonstrate the effectiveness of DW-FM on three CVaR-based contextual stochastic optimization benchmarks spanning synthetic portfolio, semi-real financial, and traffic-CVaR tasks, where DW-FM improves downstream regret over standard baselines.

AI 总结：提出精确后验分数（EPS）方法，通过闭式后验分数将线性逆问题转化为去噪问题，无需梯度或投影，在FFHQ和ImageNet上优于现有方法。

链接：https://arxiv.org/abs/2606.17048

机构：University of Oxford（牛津大学）； UIUC（伊利诺伊大学厄巴纳-香槟分校）； EverEx

作者：Abbas Mammadov, Ozgur Kara, Kaan Oktay, Iskander Azangulov, Adil Kaan Akan, Hyungjin Chung, James Matthew Rehg, Yee Whye Teh

英文摘要：Diffusion and flow-based models learn powerful data priors by training a denoiser to reverse Gaussian corruption. To use this prior to solve a linear inverse problem, one needs to sample from the posterior, but the score that the prior provides is the unconditional score, not the posterior score. Existing methods either steer a fixed pretrained denoiser with approximate measurement-matching corrections, or train a conditional restoration model that abandons the denoising structure of the prior. We derive the exact posterior score in closed form for linear Gaussian inverse problems under general Gaussian interpolants, and show that posterior sampling reduces to a denoising problem at an operator-dependent shifted pivot under an anisotropic noise covariance. We turn this identity into Exact Posterior Score (EPS), a denoising training objective that preserves the input/output structure of standard pretraining and can therefore be trained from scratch or fine-tuned from a pretrained denoiser. At inference, EPS uses the same sampler as the underlying backbone, with no likelihood gradients or projections. We evaluate EPS on five linear inverse problems across FFHQ and ImageNet, where it outperforms training-free and training-based baselines on fidelity, perceptual, and distributional metrics, while using roughly an order of magnitude fewer denoiser evaluations than gradient-based posterior samplers.

AI 总结：针对大模型微调中反向传播内存开销大、零阶优化对步长和平滑参数敏感的问题，提出统一无梯度训练、自适应调参和非欧几里得更新几何的AdaNAGED方法，并在OPT-1.3B模型上验证有效性。

链接：https://arxiv.org/abs/2606.14970

作者：Dmitriy Bystrov, Daniil Medyakov, Dmitry Bylinkin, Aleksandr Beznosikov

英文摘要：Fine-tuning large language models (LLMs) has become a central application of modern optimization, enabling pretrained models to adapt to diverse downstream tasks and domain-specific data. A major obstacle in large-scale fine-tuning is the memory overhead of backpropagation, which requires storing activations, gradients, and optimizer states. Zeroth-order (ZO) optimization offers a memory-efficient alternative, but its performance is highly sensitive to the stepsize and smoothing parameter, often requiring costly task-specific tuning. Parameter-free (PF) optimization addresses this issue by adapting algorithmic parameters without prior knowledge of problem-dependent constants. Moreover, large-scale fine-tuning can benefit from geometry-aware updates that account for the heterogeneous structure of parameter blocks, which can be modeled through methods that exploit linear minimization oracle (LMO). In this work, we study PF adaptation for LMO-based ZO optimization and introduce $\texttt{AdaNAGED}$, a method that unifies gradient-free training, adaptive tuning, and non-Euclidean update geometry. We establish convergence guarantees and validate the method on large-scale LLM fine-tuning task with $\texttt{OPT}-1.3\mathrm{B}$ model.

AI 总结：针对离线多目标优化中的分布外问题，提出DOMOO方法，通过累积风险控制、嵌套帕累托集学习和多样性驱动选择策略，在合成和真实基准上实现了收敛性和多样性的最佳平均排名。

链接：https://arxiv.org/abs/2606.15115

作者：Yiyi Zhu, Yaolin Wen, Xiang Xia, Xin An, Hanyi Si, Xiang Shu, Yangde Fu, Liang Dou, Hong Qian

英文摘要：Multi-objective optimization (MOO) has emerged as a powerful approach to solving complex optimization problems involving multiple objectives. In many practical scenarios, function evaluations are unavailable or prohibitively expensive, necessitating optimization solely based on a fixed offline dataset. In this setting, known as offline MOO, the goal is to find out the Pareto set without access to the true objective functions. This setting suffers from the out-of-distribution (OOD) issue, where the surrogate model is not accurate for unseen designs. Due to the OOD issue, surrogate errors may cause the optimizer to select solutions that do not lie on the true Pareto front and are biased toward its extremes. To address this, this paper proposes Diversity-driven Offline Multi-Objective Optimization (DOMOO), which aims to find out a diverse and high-quality set of solutions. First, DOMOO incorporates an accumulative risk control module that estimates the potential risk of candidate solutions and alleviates the OOD issue between the training data and the generated solutions. In addition, a nested Pareto set learning (PSL) strategy is proposed to jointly learn preference and PSL parameters, then optimize them, enabling adaptation to diverse Pareto front geometries. To further enhance solution quality, we design a diversity-driven selection strategy that extracts a representative and well-distributed set of final solutions. To achieve this diversity-driven selection strategy, we propose $\text{IGD}_\text{offline}$, a tailored indicator for the offline setting that considers both diversity and convergence, and avoids the bias of hypervolume indicator. Extensive experiments on synthetic and real-world benchmarks show that DOMOO achieves the best average rank across tasks in both convergence and diversity among the compared methods.

AI 总结：提出统一梯度算法训练两层神经网络，在多项式时间内学习高斯单指标模型，样本复杂度匹配SQ下界，并扩展到稀疏情形。

链接：https://arxiv.org/abs/2606.15219

作者：Siyu Chen, Beining Wu, Miao Lu, Zhuoran Yang, Tianhao Wang

英文摘要：In this work, we tackle the following question: Can neural networks trained with gradient-based methods achieve the optimal computational-statistical tradeoff in learning Gaussian single-index models? Prior research has shown that any polynomial-time algorithm under the statistical query (SQ) framework requires $\Omega(d^{s^\star/2}\lor d)$ samples, where $s^\star$ is the generative exponent representing the intrinsic difficulty of learning the underlying model. However, it remains unknown whether neural networks can achieve this sample complexity. Inspired by prior techniques such as label transformation and landscape smoothing for learning single-index models, we propose a unified gradient-based algorithm for training a two-layer neural network in polynomial time. Our method is adaptable to a variety of loss and activation functions, covering a broad class of existing approaches. We show that our algorithm learns a feature representation that strongly aligns with the unknown signal $\theta^\star$, with sample complexity $\widetilde{O} (d^{s^\star/2} \lor d)$, matching the SQ lower bound up to a polylogarithmic factor for all generative exponents $s^\star\geq 1$. Furthermore, we extend our approach to the setting where $\theta^\star$ is $k$-sparse for $k = o(\sqrt{d})$ by introducing a novel weight perturbation technique that leverages the sparsity structure. We derive a corresponding SQ lower bound of order $\widetilde{\Omega}(k^{s^\star})$, matched by our method up to a polylogarithmic factor. Our framework, especially the weight perturbation technique, is of independent interest, and suggests potential gradient-based solutions to other problems such as sparse tensor PCA.

AI 总结：本文通过理论分析解决关于 Schatten-∞ 优化器有效性的矛盾观察，发现结论取决于数据维度：在低维场景（包括 Chinchilla 缩放）下，较小的 Schatten-p 几何更优，并基于 SODA 框架为 p>2 提出新的噪声鲁棒加速结果。

链接：https://arxiv.org/abs/2606.15268

作者：Thomas Pethick

英文摘要：Schatten-$\infty$ based optimizers such as Muon have shown promising empirical performance, but there remains seemingly conflicting observations regarding whether they are beneficial. We resolve this conflict by showing that the conclusion is regime dependent. Even when the objective is smooth in the Schatten-$\infty$ geometry, smaller Schatten-$p$ geometries can be optimal, specifically in the low-dimensional regime, which we show includes Chinchilla scaling. This conclusion follows from a new noise-robust acceleration result for the SODA framework for $p>2$. The same analysis explains why Muon-like methods do not require warmup, why they naturally favor large batches, and yields a batch size scaling rule for arbitrary $p$.

AI 总结：本文通过过度训练的视角形式化RLVR中的多样性崩溃，发现标准训练中大部分更新是过度训练，并提出贝叶斯边界门控（BBG）方法，通过估计每个问题对推理边界的边际贡献来优化，提升多个基准上的Pass@k。

链接：https://arxiv.org/abs/2606.15455

机构：Sydney AI Centre, The University of Sydney（悉尼大学悉尼人工智能中心）； Southeast University（东南大学）； Microsoft（微软）； Data61, CSIRO（澳大利亚联邦科学与工业研究组织Data61）； Chongqing University（重庆大学）； Nanyang Technological University（南洋理工大学）

作者：Suqin Yuan, Jinkun Chen, Jiyang Zheng, Muyang Li, Lei Feng, Dadong Wang, Tao Xiang, Tongliang Liu, Bo An

英文摘要：Reinforcement learning with verifiable rewards (RLVR) has become a key approach for enhancing the reasoning abilities of large language models. However, RLVR often suffers from \emph{diversity collapse}: Pass@$1$ improves while high-$k$ Pass@$k$ degrades, which is viewed as a narrowing of the model's reasoning boundary. We formalize this diversity collapse through the lens of \emph{overtraining}: once a problem's contribution to the reference metric has effectively saturated, further updates no longer expand what the model can solve but still concentrate probability mass on the trajectories favored by on-policy sampling. Under a standard setup with few rollouts per problem, even a single observed success places a problem in a nearly saturated regime for high-$k$ Pass@$k$, so most updates in standard RLVR are overtraining from the boundary perspective. This perspective also suggests a reading of whether RLVR can expand the model's reasoning abilities beyond the base model: since RLVR is structurally biased against high-$k$ Pass@$k$, its aggregate decline does not by itself mean that no new reasoning gains occurred. Interventionally, restricting updates to problems with zero observed success lifts Pass@$256$ above the base model on difficult benchmarks; observationally, a non-trivial fraction of initially unsolvable problems become solvable during standard RLVR training. Building on these findings, we propose \emph{Bayesian Boundary Gating} (BBG), which redirects optimization away from overtraining by estimating each problem's marginal contribution to the reasoning boundary. Across multiple reasoning benchmarks, BBG improves average Pass@$k$ across a wide range of $k$.

AI 总结：提出分岔理论框架，通过将训练动力学分解为法向和切向分量，证明稳定性边缘训练源于法向的翻转分岔，并收敛到最小化流形。

链接：https://arxiv.org/abs/2606.15551

作者：Eric Gan

英文摘要： The Edge of Stability (EoS) phenomenon, where gradient descent operates with sharpness exceeding the classical convergence threshold yet the loss decreases over long timescales, is ubiquitous in modern deep learning but remains poorly understood in realistic settings. Prior rigorous analyses have been largely confined to scalar or low-dimensional losses with specific structural forms. In this work, we develop a bifurcation theory framework for gradient descent on the edge of stability that applies directly to overparameterized neural networks. By decomposing the training dynamics into components normal and tangent to the manifold of minimizers, we show that stable EoS training arises from a flip bifurcation in the normal direction, governed by the sign of the first Lyapunov coefficient, while the tangent dynamics drift toward regions of decreasing sharpness. Under mild spectral and geometric assumptions on the loss landscape, we prove convergence to the minimizing manifold when training at the EoS threshold. As a corollary, we recover and unify prior results: we show that the product-stability condition of Gan (2026) is an instance of our framework.

AI 总结：通过决策论将测试时训练视为核机制下的隐式贝叶斯推断，揭示了更新步长和子空间选择对性能的影响，并提出了自适应策略、PAC-Bayes保证和最优子空间选择规则。

链接：https://arxiv.org/abs/2606.15569

作者：Tomoya Wakayama

英文摘要：Test-time training (TTT) adapts a pretrained model to each prompt via parameter updates, improving accuracy under pretraining-to-test distribution shifts. Yet, its performance often suffers from instability and sensitivity to hyperparameters such as update steps and subspace. We explain this behavior through a decision-theoretic lens, treating TTT as implicit Bayesian inference in the kernel regime. Under a Gaussian process benchmark, we show that TTT reduces prediction error when updates are spectrally matched to the prompt's signal-to-noise ratio and aligned with query-relevant eigen-directions. This perspective underpins the following results: (1) we show when fixed update steps and subspaces fail under distribution shifts, motivating adaptive strategies; (2) we prove that selecting update steps via prompt evidence admits a PAC-Bayes guarantee against overfitting; and (3) we characterize the Bayes-optimal update subspace under a linear-Gaussian correction model, yielding a scoring rule for selecting Transformer blocks and heads. Our theory helps explain the empirical instability of TTT, taking a step toward principled guidance for when, how far, and which directions to adapt.

AI 总结：针对异质噪声数据，提出多保真度SINDy方法，通过加权回归融合集成SINDy和弱SINDy，从不同保真度测量中稀疏识别非线性动力系统，理论证明加权策略的统计合理性，在常微分和偏微分方程基准系统及双摆预测中验证了其抑制异方差噪声、利用低成本低质量数据提升模型恢复的效果。

链接：https://arxiv.org/abs/2606.15690

机构：Politecnico di Milano（米兰理工大学）； University of Washington（华盛顿大学）

作者：Filippo Zacchei, Ana Larrañaga, Attilio Frangi, Andrea Manzoni, Steven L. Brunton

英文摘要：Data from simulations and experiments are rarely noise-free and often exhibit heterogeneous levels of fidelity. Measurement uncertainty may vary across repeated observations, sensing devices, or even within a single experiment. This work addresses the problem of discovering nonlinear dynamical systems from such inhomogeneous data. We extend the Sparse Identification of Nonlinear Dynamical Systems (SINDy) framework to account for variable noise levels by combining Ensemble SINDy and Weak SINDy within a weighted regression formulation derived from generalized least squares. A statistical justification for the weighting strategy is also provided. The methodology is validated on several benchmark systems, including ordinary and partial differential equations. In addition, we show the benefit of multi-fidelity integration for forecasting the dynamics of a double pendulum system. The results confirm that the proposed approach mitigates the adverse effects of heteroscedastic noise and that repeated, low-cost, low-quality measurements can improve model recovery, in some cases matching or outperforming reconstructions obtained using only high-fidelity data.

AI 总结：提出布朗核梯子（BKL）递归层次函数空间，通过布朗核积分构造，证明其为准Banach空间且具有深度相关Hölder正则性，为深度学习的组合表示提供可解析框架。

链接：https://arxiv.org/abs/2606.15812

作者：Mahdi Mohammadigohari, Giuseppe Di Fatta, Giuseppe Nicosia, Panos M Pardalos

英文摘要：Constructing mathematically tractable function spaces that capture hierarchical compositional representations remains a central challenge in statistical learning theory. We introduce Brownian kernel ladders (BKLs), a recursively defined hierarchy of integral reproducing kernel Hilbert spaces generated through Brownian-kernel integral constructions. Starting from linear functionals, each layer is obtained by integrating Brownian kernels over probability measures supported on subsets of the previous layer, yielding a recursive function-space model in which depth is encoded directly through the hierarchy. Based on this framework, we define canonical BKL spaces together with an associated complexity functional. We establish several analytical and statistical properties of these spaces. In particular, we show that BKL spaces form quasi-Banach spaces, satisfy depth-dependent Hölder regularity estimates, and exhibit strict monotonicity with respect to depth. We further prove existence results for regularized empirical risk minimization and derive Gaussian complexity bounds that remain uniformly controlled with respect to both the ambient dimension and the hierarchy depth. A key ingredient of the analysis is a combinatorial proof technique based on recursive subset decompositions and Brownian-kernel threshold representations. These estimates yield excess-risk guarantees of near-parametric order for regularized empirical risk minimization over BKL spaces. Our results provide a mathematically tractable hierarchical function-space framework for studying compositional representations in deep learning.

AI 总结：针对大规模数据中嵌套双有限和结构的非凸优化，提出SILAGE算法，通过利用双和结构避免全局全梯度刷新，仅需O(n)内存，并基于组间和组内异质性实现自适应收敛分析。

链接：https://arxiv.org/abs/2606.15832

作者：Igor Sokolov, Laurent Condat, Peter Richtárik

英文摘要：Empirical risk minimization on massive datasets naturally exhibits a nested double finite-sum structure, where $N=nm$ total samples are logically or physically partitioned into $n$ blocks of size $m$ (e.g., in pooled data silos, out-of-core learning, or deliberate stratification). While variance-reduced methods achieve optimal oracle complexities for nonconvex objectives, they suffer from severe scaling bottlenecks in this centralized regime. Recursive estimators, such as PAGE, require periodic global full-gradient refreshes over all $nm$ samples, which are computationally expensive. Conversely, single-loop methods, such as SILVER, avoid such refreshes but require an impractical $\mathcal{O}(nm)$ memory footprint to store a control variate for every sample. In this paper, we propose SILAGE, a variance-reduced algorithm that addresses this trade-off. By actively exploiting the double-sum structure, SILAGE eliminates periodic global full-gradient refreshes over all $nm$ components (evaluating at most one local group gradient per iteration) while requiring only $\mathcal{O}(n)$ memory. Furthermore, we provide a tight convergence analysis that avoids pessimistic worst-case Lipschitz constants. Instead, SILAGE's complexity natively adapts to the underlying data geometry via nested functional similarities: across-group ($\delta_1$) and within-group ($\delta_2$) heterogeneity. Our results improve existing state-of-the-art bounds in several practically relevant regimes.

AI 总结：本文通过信息论框架，证明在正交约束下，联合近似比单独近似需要更少的描述长度，揭示了共享表示在组合架构中的效率优势。

链接：https://arxiv.org/abs/2606.16028

机构：Johann Radon Institute of Computational and Applied Mathematics, Austrian Academy of Sciences（奥地利科学院约翰·拉东计算与应用数学研究所）； Faculty of Mathematics, University of Vienna（维也纳大学数学学院）

作者：Thomas Dittrich, Oliver Potocki, Philipp Grohs

英文摘要：Modern deep learning architectures are increasingly multi-task and multi-modal, using a pretrained foundation model combined with task-specific, fine-tuned models. Empirically, exploiting similarity across different problems, instead of solving them individually, can significantly improve overall performance. While the generalization and sample complexity properties of multitask learning have been widely studied, the parametric complexity of joint approximation in comparison to separate approximation remains less well understood. The question is particularly relevant in modern deep learning, where models are increasingly required to satisfy structural constraints such as equivariance, conservation laws, or orthogonality. We prove lower and upper bounds on the description-length for separate and joint approximation classes, respectively, in uniform norm. We build a class of orthogonal functions by composing a shared hard feature, realized by a Rademacher-Haar wavelet series, with Sawtooth-Walsh readouts to enforce orthogonality of output coordinates. The dyadic tree structure of the Rademacher-Haar wavelet concentrates the approximation hardness in the common feature component, while the readouts act as task-specific heads. Using an information-theoretic framework, we obtain a sharp gap between the optimal approximation rates achievable by joint and separate coding. Finally, we realize this separation in a neural network model using Heaviside activations via reduction to triangle-wave approximation. Our results show that even under an orthogonality constraint joint approximation requires strictly fewer bits in compositional architectures, provided the tasks share a latent hard feature. This provides theoretical insight into the description-length-efficiency of compositional multi-output architectures and clarifies how neural networks can retain expressivity under geometric constraints.

AI 总结：针对非对数凹分布采样中随机梯度高方差问题，提出统一分析动量、STORM和PAGE等方差缩减方法，证明其在相对Fisher信息和非平方总变差距离下的改进收敛率，并扩展至基于得分的生成先验逆问题求解。

链接：https://arxiv.org/abs/2606.16257

作者：M. Berk Sahin, Ahmet Ege Tanriverdi, Behzad Sharif, Abolfazl Hashemi

英文摘要：Sampling from high-dimensional, non-log-concave distributions with unnormalized densities is a fundamental challenge in machine learning, particularly when the exact gradient of the potential is unavailable and must be approximated via stochastic gradients that exhibit high variance under a fixed budget of gradient computations per iteration. Although variance reduction techniques such as SGD with momentum, STORM, and PAGE have demonstrated improved convergence properties in non-convex optimization, their implications for sampling from non-log-concave distributions remain largely unexplored. In this work, we develop the first unified analysis of these estimators for sampling from non-log-concave distributions. We establish improved non-asymptotic convergence rates in $\varepsilon$-relative Fisher information and, under a Poincaré inequality assumption, in squared total variation distance, and further prove weak convergence to the target distribution. We extend our analysis to solving inverse problems with score-based generative priors. We empirically validate our theory and demonstrate that, under a fixed gradient computations per iteration, variance-reduction techniques consistently improve sample quality in two standard imaging applications.

AI 总结：提出GENIE优化器，通过一步泛化比率（OSGR）动态均衡参数更新，抑制虚假相关，促进域不变特征学习，在域泛化任务中超越现有优化器。

链接：https://arxiv.org/abs/2606.16301

作者：Sumin Cho, Dongwon Kim, Kwangsu Kim

英文摘要： Domain Generalization (DG) aims to train models that generalize to unseen target domains but often overfit to domain-specific features, known as undesired correlations. Gradient-based DG methods typically guide gradients in a dominant direction but often inadvertently reinforce spurious correlations. Recent work has employed dropout to regularize overconfident parameters, but has not explicitly adjusted gradient alignment or ensured balanced parameter updates. We propose GENIE (Generalization-ENhancing Iterative Equalizer), a novel optimizer that leverages the One-Step Generalization Ratio (OSGR) to quantify each parameter's contribution to loss reduction and assess gradient alignment. By dynamically equalizing OSGR via a preconditioning factor, GENIE prevents a small subset of parameters from dominating optimization, thereby promoting domain-invariant feature learning. Theoretically, GENIE balances convergence contribution and gradient alignment among parameters, achieving higher OSGR while retaining SGD's convergence rate. Empirically, it outperforms existing optimizers and enhances performance when integrated with various DG and single-DG methods.

AI 总结：将Menon和Yu的实深度线性网络熵公式推广到复和四元数情形，得到统一公式。

链接：https://arxiv.org/abs/2606.16579

机构：CINVESTAV-IPN（墨西哥国立理工学院高级研究中心）； Brown University（布朗大学）

作者：Luis Contreras, Marco Nahas, Tejas Kotwal

英文摘要：We extend the entropy formula of Menon and Yu for the real Deep Linear Network (DLN) to its complex and quaternionic analogues, obtaining a unified formula for DLNs over $\mathbb{R}$, $\mathbb{C}$, and $\mathbb{H}$.

AI 总结：提出基于LangGraph的有状态ReAct智能体，通过持久化状态和固定大小对话窗口，将自主实验的令牌成本从O(n²)降至O(1)，在超参数调优和代码优化任务中分别减少90%和52%的令牌消耗。

链接：https://arxiv.org/abs/2606.14945

作者：Faramarz Jabbarvaziri

英文摘要：The autoresearch pattern enables autonomous experimentation by having a large language model (LLM) iteratively modify code to optimize a target metric. Its stateless design, however, reconstructs experimental context from scratch at every iteration, incurring $O(n)$ token cost per iteration and $O(n^{2})$ total. This work reformulates the pattern as a stateful ReAct agent using LangGraph, where typed persistent state carries experimental history across iterations via a tool-calling interface. Two benchmarks are evaluated: hyperparameter tuning (15 iterations, small per-iteration observations) and code performance optimization (40 iterations, large per-iteration observations containing full source code and benchmark results). On hyperparameter tuning, the stateful agent consumes 90\% fewer tokens (2{,}492 vs.\ 24{,}465). On code optimization, the stateful agent consumes 52\% fewer tokens (627K vs.\ 1{,}275K) while achieving comparable optimization quality on both tasks. The token reduction is structural: the stateless agent re-reads the full history at $O(n)$ cost per iteration, while the stateful agent operates within a fixed-size conversation window at $O(1)$ cost. This paper describes the architecture in sufficient detail for practitioners to implement a stateful autoresearch agent for their own workflows.

AI 总结：针对长上下文大模型推理中KV缓存压缩问题，提出PolyKV框架，通过层级别信号为每层选择合适压缩策略并分配非均匀缓存预算，实验表明在固定预算下显著恢复性能差距。

链接：https://arxiv.org/abs/2606.15157

机构：King Abdullah University of Science and Technology（阿卜杜拉国王科技大学）

作者：Chao Fei, Panos Kalnis

英文摘要：KV cache compression is essential for reducing the memory cost of long-context large language model inference. Existing approaches, however, typically apply a single compression policy and a uniform cache budget across all transformer layers. This uniform design ignores the fact that different layers can play different roles during prefill and decoding, and may therefore require different eviction strategies and cache capacities. We present PolyKV, a layer-wise KV cache optimization framework that considers design space with method selection and budget allocation. PolyKV routes each layer to a suitable KV compression policy based on layer-level signals, while assigning non-uniform budgets under a fixed total budget. This formulation enables heterogeneous compositions of existing KV cache methods. Experiments on LLaMA-3.1-8B and Qwen3-8B show that, under the same 512-token average KV budget, PolyKV recovers 54.5% and 25.7% of the LongBench performance gap between the strongest single-policy baseline and FullKV, respectively. Across 128-1024 budget sweep, PolyKV consistently improves over the strongest baseline by 1.7%-6.4%, corresponding to 40.0%-54.5% recovery of the FullKV gap.

AI 总结：提出M-CTX框架，将空间上下文构建转化为空间数据库查询，通过索引加速实现226倍加速，解决轨迹预测中上下文构建的系统瓶颈。

链接：https://arxiv.org/abs/2606.15244

机构：Harbin Engineering University（哈尔滨工程大学）； Wuhan University of Technology（武汉理工大学）； University of Chinese Academy of Sciences（中国科学院大学）； Alibaba Group（阿里巴巴集团）

作者：Kun Ma, Qilong Han, Chengjing Song, Jingzheng Yao, Xiao Han, Yuee Zhou, Changmao Wu

英文摘要：Modern trajectory predictors increasingly condition on external spatial context, such as map geometry, signed distance fields (SDFs), and nearby moving agents. While this context improves prediction quality, constructing it for every training anchor has become a hidden systems bottleneck. In a representative maritime AIS pipeline, spatial context construction requires roughly 17 CPU-days for a 5.48M-anchor corpus, dominating the cost of the downstream predictor. We present M-CTX, an exact and scalable spatial context-retrieval framework for trajectory analytics. M-CTX recasts context construction as an ingest-once, query-many spatial database workload and replaces three brute-force stages -- OSM range retrieval, SDF computation, and moving-vessel neighbour lookup -- with composable, index-backed operators. Its learned range-index backend, BR-LZ, provides recall-complete MBR-overlap range retrieval and reduces candidate amplification by 1.1x--2.7x relative to global-expansion one-curve baselines. Across four maritime regions, eight baseline systems, synthetic workloads with up to 40M spatial features, and 10^7-record AIS streams, M-CTX reproduces the reference context exactly. On the 5.48M-anchor corpus, it reduces context construction from about 17 CPU-days to 1.8 hours, a measured 226x end-to-end speed-up. An optional storage mode further compresses SDF context by 64x with only a 0.04 m ADE change. These results establish exact spatial context retrieval as a first-class database problem in modern trajectory analytics. Code and datasets are publicly available at this https URL.

AI 总结：提出Drift-RAE方法，通过漂移范式在表示自编码器潜空间中蒸馏预训练流模型，解决各向异性和大曲率问题，在ImageNet 256上仅用10k步达到1.77 FID。

链接：https://arxiv.org/abs/2606.15553

机构：Tsinghua University（清华大学）； Ant Group（蚂蚁集团）； CUHK（香港中文大学）

作者：Jiawei Zhang, Mengfei Xia, Gen Li, Yuantao Gu

英文摘要：Representation Autoencoders (RAEs) have improved diffusion and flow models by semantically richer latent space owing to the strongly label-wise clustered DINO features in the pretrained encoders. Yet in the distillation stage, the severe anisotropy and large curvatures caused by the rich semantic representations would hinder the convergence and performance, making the trajectory-based distillation unstable. In this work, we argue that the RAE latent space is compatible with distillation via the newly proposed Drifting Models. We first quantitatively study the curvatures and isotropy statistics across different autoencoders, and theoretically reveal that Drifting Model itself is highly likely to fail on extremely scattered spaces like reconstruction-based VAEs. These motivate us to apply the drifting paradigm directly to representation autoencoders. Our proposed method, Drift-RAE, distills pretrained flow models in RAE latent spaces using Drifting, together with insightful modifications that improve training stability by thereotically aligning drifting fields with other frameworks. Regarding the experimental evidences, we achieve 1.77 FID on ImageNet 256 dataset using only 10k distillation steps, surpassing state-of-the-art RAE distillation methods and appearing comparative with the original Drifting Model without requiring an auxiliary MAE feature extractor. The code will be made publicly available.

AI 总结：针对DiT-MoE中跨时间步的冗余计算，提出基于专家分支级别的细粒度缓存框架MoECa，实现分支级特征复用，并引入专家感知自适应控制和同步缓存更新，在多个模型上取得高达2.83倍加速且质量损失极小。

链接：https://arxiv.org/abs/2606.15615

机构：School of Computer Science, Peking University（北京大学计算机科学学院）； School of Software Engineering, University of Electronic Science and Technology of China（电子科技大学软件工程学院）

作者：Maoliang Li, Haojing Chen, Jiayu Chen, Zihao Zheng, Xinhao Sun, Hailong Zou, Xiang Chen

英文摘要：Diffusion Transformers with Mixture-of-Experts (DiT-MoE) improve model capacity under sparse activation, but diffusion inference is still bottlenecked by redundant computation across timesteps. Existing caching methods mainly operate at the token level, which becomes suboptimal in DiT-MoE because each token update is internally decomposed into multiple routed expert branches. Our analysis shows that cross-timestep redundancy in DiT-MoE is better characterized at the expert-branch level than at the whole-token level. Based on this observation, we propose MoECa, a fine-grained caching framework that performs branch-level feature reuse across timesteps. MoECa further introduces expert-aware adaptive control and synchronized cache updates across MoE and attention paths to maintain stable intermediate states. Experiments on multiple DiT-MoE models show that MoECa consistently achieves a better speed-quality trade-off than prior caching methods, with up to 2.83$\times$ inference speedup and minimal quality degradation.

AI 总结：提出MosaicQuant，通过将权重矩阵量化为密集4位基分量和稀疏4位残差分量，结合ZipperEngine融合稀疏块计算，实现统一4位推理，在LLaMA3和Qwen3上保持近FP16精度并加速1.24倍。

链接：https://arxiv.org/abs/2606.15652

机构：HKUST（香港科技大学）； EPFL（瑞士联邦理工学院洛桑）； MetaX Integrated Circuits Co., Ltd（MetaX集成电路有限公司）

作者：Yangjia Hu, Haodong Wang, Zicong Hong, Qianli Liu, Quanxin Shou, Jian Lin, Song Guo, Xiaowei Shen, Xiangjun Huang, Dian Wang, Jian Yang

英文摘要： 4-bit quantization significantly reduces the memory footprint and accelerates the inference of large language models (LLMs). However, its limited bit-width representation struggles to faithfully capture both dense common values (\emph{inliers}) and rare large-magnitude values (\emph{outliers}), causing substantial accuracy degradation. Existing mixed-precision methods mitigate this by retaining outliers in high precision, but at the cost of breaking the uniformity of low-bit execution, introducing precision conversion and extra data movement that undermine practical speedup. We propose \textbf{MosaicQuant}, a unified 4-bit LLM quantization paradigm built on a novel principle of \emph{inlier--outlier disaggregation}. Rather than elevating outlier precision, MosaicQuant quantizes the full weight matrix into a dense 4-bit base component, where inliers are captured faithfully while outlier are inevitably quantized. A sparse 4-bit residual component is then introduced to compensate for these quantization errors, selectively targeting the most error-critical weight blocks where output distortion is shown to be concentrated. However, a unified representation alone is insufficient, as naïvely executing the sparse residual as a separate kernel still breaks the unified low-bit inference pipeline. To bridge this gap, we introduce \textbf{ZipperEngine}, which fuses sparse block computation into the dense 4-bit GEMM kernel via an overlapped pipeline, unifying not only the representation but also the execution into a single coherent low-bit inference pipeline. Extensive experiments on LLaMA3 and Qwen3 demonstrate that MosaicQuant preserves near-FP16 accuracy while achieving up to $1.24\times$ speedup over the W16A16 baseline.

AI 总结：针对大推理模型在低比特量化（W4A4KV4）下推理精度严重下降的问题，提出ReQAT框架，通过迹对齐QAT、选择性熵最小化和量化友好初始化，恢复并超越BF16微调精度，实现最高3.9倍吞吐加速。

链接：https://arxiv.org/abs/2606.15682

机构：Hanyang University（汉阳大学）； Samsung Advanced Institute of Technology（三星综合技术院）

作者：Janghwan Lee, Sihwa Lee, Jinseok Kim, Yongjik Kim, Jieun Lim, Jinwook Oh, Jungwook Choi

英文摘要：Large Reasoning Models (LRMs) achieve strong problem-solving through long chain-of-thought, but their deployment is constrained by the high cost of full-precision inference and growing KV cache footprints. Microscaled FP4 formats enable efficient FP4 deployment; however, fully quantizing weights, activations, and KV caches (W4A4KV4) causes severe reasoning degradation that existing PTQ and QAT fail to recover. We identify that FP4 failures concentrate on low-entropy tokens--precise symbolic commitments such as digits and operators--where quantization noise inflates sampling errors that cascade through reasoning traces. Based on this insight, we propose ReQAT, a reasoning-centric FP4 training framework with three components: (i) Trace-Aligned QAT (TAQ), which revisits identical reasoning traces to focus updates on critical low-entropy decisions; (ii) Selective Entropy Minimization (SEM), which reinforces confidence at low-entropy positions; and (iii) Q-FIT, a quantization-friendly initialization that jointly calibrates RoPE-consistent KV cache transformations to stabilize QAT. Under the same training budget, ReQAT not only recovers but surpasses BF16 fine-tuning accuracy, while delivering up to 3.9x throughput speedup on NVIDIA DGX Spark and 3.1x on B200.

AI 总结：提出一次性MoE专家剪枝的统一公式，基于路由频率、门控权重和激活强度三个因素，推导出任务无关剪枝应使用基于激活的准则，任务特定剪枝可保留路由频率和门控信息，并据此提出两种新准则MAN和MSAN，在多个模型和基准上取得最优性能。

链接：https://arxiv.org/abs/2606.15716

机构：Zhejiang University（浙江大学）； Westlake University（西湖大学）； Mohamed bin Zayed University of Artificial Intelligence（穆罕默德·本·扎耶德人工智能大学）； Carnegie Mellon University（卡内基梅隆大学）

作者：Zongfang Liu, Jinghui Zhang, Zijian Ma, Guangyi Chen, Xin Yuan

英文摘要：Mixture-of-Experts (MoE) language models reduce per-token computation through sparse expert activation, yet deployment still requires storing the full expert pool, making one-shot expert pruning a practical approach for reducing memory usage. Although effective, existing criteria are largely heuristic, and no single criterion is universally optimal. Thus, establishing a principle for selecting pruning criteria suited to different deployment objectives remains an important yet largely underexplored problem in one-shot expert pruning. To this end, we introduce a unified formulation for one-shot MoE expert pruning organized around three factors: routing frequency, gate weighting, and activation strength. The formulation yields a criteria selection principle: task-agnostic pruning should favor routed-token-averaged, gate-free activation-based criteria, whereas task-specific pruning can benefit from retaining routing-frequency and gate-weight information. Beyond this principle, the formulation also provides a systematic view of existing heuristic criteria and gives rise to two new task-agnostic criteria, Mean Activation Norm (MAN) and Mean Squared Activation Norm (MSAN). Across four representative MoE models and 16 diverse benchmarks, MAN and MSAN are consistently strong in the task-agnostic setting, obtain the top-two average ranks, and improve average performance by up to 8.8 points over the strongest baseline.

AI 总结：针对多轮智能体在线策略蒸馏中错误累积导致教师监督失效的问题，提出混合教师和学生生成回合的Guided-OPD算法，通过课程式衰减教师干预概率，在ALFWorld等任务上平均提升21.1%得分和25.5%成功率。

链接：https://arxiv.org/abs/2606.15912

机构：Foundation Model Research Center, Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所基础模型研究中心）； School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）； Large Language Model Department, Tencent（腾讯大语言模型部）； University of Science and Technology of China（中国科学技术大学）； Zhejiang University（浙江大学）； National University of Singapore（新加坡国立大学）； Wuhan AI Research（武汉人工智能研究院）

作者：Gengsheng Li, Mao Zheng, Mingyang Song, Ruiqi Liu, Tianyu Yang, Jie Sun, Qiyong Zhong, Haiyun Guo, Junfeng Fang, Dan Zhang, Jinqiao Wang

英文摘要： Multi-turn agents that plan, invoke tools, and interact with environments offer a promising paradigm for solving complex tasks, yet their capabilities typically rely on very large models whose inference cost is prohibitive in this http URL -Policy Distillation (OPD) is a natural recipe for transferring such capabilities to smaller students, but we find that it suffers a characteristic failure mode in this setting: small student errors compound across turns and push the trajectory out of the teacher's familiar state distribution, so the teacher's supervision becomes least reliable precisely where the student needs it this http URL propose Guided On-Policy Distillation (Guided-OPD), a simple yet effective algorithm that mixes teacher- and student-generated turns within each rollout and schedules the teacher's intervention probability along a curriculum that decays to this http URL guidance keeps early trajectories close to the teacher distribution and is then gradually withdrawn to recover the purely on-policy regime used at this http URL ALFWorld, ScienceWorld, and WebShop, distilling Qwen3 students from a Qwen3-30B-A3B teacher, Guided-OPD improves Score by 21.1\% and Success Rate by 25.5\% over vanilla OPD on average, with larger gains on smaller students.

AI 总结：本文介绍Mojo语言，通过MLIR编译和确定性内核设计，解决量化金融中Python到C++的性能差距与数值不一致问题，在金融AI工作负载上实现20-180倍加速。

链接：https://arxiv.org/abs/2606.16059

机构：Data Science and Artificial Intelligence Innovation Laboratory, School of Engineering and Computer Science, Baylor University（贝勒大学工程与计算机科学学院数据科学与人工智能创新实验室）

作者：Henry Han

英文摘要：For thirty years, quantitative finance has paid a costly two-language tax: models researched in Python are rewritten in C++ for production, often introducing numerical discrepancies. GPU-accelerated deep learning exacerbates this problem, as nondeterministic floating-point reductions can produce drift in long backtests, challenging regulatory reproducibility and auditability expectations. This article surveys Mojo, Modular's 2026 Python-like systems language, as a structural response for capital markets engineering. While closing the Python-to-C++ performance gap, Mojo uniquely combines native interoperability with the low-level systems control required to construct bit-exact deterministic kernels. Its MLIR compilation infrastructure further allows a single codebase to target scalar, SIMD, multicore, and GPU execution, reducing the translation bottleneck between research and production. We benchmark four core financial AI workloads: Monte Carlo option pricing, LLM sentiment inference, multi-asset backtesting, and portfolio Value at Risk. On Apple Silicon, Mojo demonstrates 20x to 180x speedups over pure Python on directly measured kernels; larger-scale GPU workload results are projections calibrated from published benchmarks. Alongside transparent performance data, we introduce mojo-deterministic, an open-source library of reproducible reduction kernels, and provide a candid assessment of the problems Mojo does and does not yet solve.

AI 总结：提出一种轻量级硬件感知神经架构搜索方法，生成可在超低功耗微控制器上运行的微型CNN，在保持分类精度的同时降低搜索成本。

链接：https://arxiv.org/abs/2606.16290

机构：University of Genoa（热那亚大学）； Scuola Superiore Sant’Anna（圣安娜高等研究学院）

作者：Andrea Mattia Garavagno, Edoardo Ragusa, Antonio Frisoli, Paolo Gastaldo

英文摘要：Hardware-aware neural architecture search (HW-NAS) allows the integration of Convolutional Neural Networks (CNNs) in microcontrollers devices by automatically designing neural architectures that can fit prearranged hardware constraints. However, state-of-the-art HW-NAS target high-performance microcontrollers, whose power consumption does not meet sensing nodes requirements. This work presents a HW-NAS generating tiny CNNs that can run on ultra-low-power microcontrollers, featuring a lightweight search procedure enabling its execution even on embedded devices. Empirical results on three well-known benchmarks for tiny computer vision proved that the proposed HW-NAS was able to generate tiny CNNs while preserving state-of-the-art classification accuracy.

AI 总结：提出VeriAttn，通过将注意力计算卸载到GPU并由TEE验证，结合两阶段流水线和分区策略，显著降低TEE计算和通信开销，实现LLM推理加速。

链接：https://arxiv.org/abs/2606.16352

机构：Nanyang Technological University（南洋理工大学）； Zero Gravity Labs（零重力实验室）

作者：Ziqun Chen, Ming Wu, Michael Heinrich, Jason Zeng, Huiying Lan, Tianwei Zhang, Rui Tan

英文摘要：Computation integrity of remote large language model (LLM) serving can be questionable. For conventional deep neural networks (DNNs), the existing TEE-shielded DNN partitioning (TSDP) approach uses Trusted Execution Environment (TEE) to compute non-linear components and verify the integrity of linear components offloaded to an untrusted GPU. However, directly applying TSDP to Transformer-based LLMs incurs significant TEE computation and TEE-GPU communication overhead. This paper presents Communication-efficient TEE-GPU Attention (\textsc{VeriAttn}) for accelerating verifiable LLM inference. \textsc{VeriAttn} offloads both linear and non-linear computations of attention to the GPU, while TEE performs verification. Moreover, for prefill, \textsc{VeriAttn} uses a two-level pipeline to overlap data movement, TEE pre-/post-processing, and GPU computation. For decoding, when the key-value cache exceeds available GPU memory, \textsc{VeriAttn} partitions attention across TEE and GPU to reduce repeated key-value transfers. Evaluation on an Intel TDX platform shows that \textsc{VeriAttn} achieves 2.60-3.38$\times$ and 3.86-5.42$\times$ acceleration over TSDP for 6k-token prompts and 10k-token outputs during prefill and decoding, respectively.

AI 总结：提出一种基于子空间混合的压缩方法，在低带宽分布式训练中实现超过95%的通信压缩，支持百亿参数模型在100K上下文长度下高效训练。

链接：https://arxiv.org/abs/2606.16384

机构：Pluralis Research

作者：Sameera Ramasinghe, Ajanthan Thalaiyasingam, Hadi Mohaghegh Dolatabadi, Gil Avraham, Violetta Shevchenko, Yan Zuo, Chamin Hewa Koneputugodage, Alexander Long

英文摘要：Pretraining language models with extended context windows enhances their ability to leverage rich information during generation. Existing methods split input sequences into chunks, broadcast them across multiple devices, and compute attention block by block which incurs significant communication overhead. While feasible in high-speed clusters, these methods are impractical for decentralized training over low-bandwidth connections. We propose a compression method for communication-efficient context parallelism in decentralized settings, achieving a remarkable compression rate of over 95\% with negligible overhead and no loss in convergence. Our key insight is to exploit the intrinsic low-rank structure of activation outputs by dynamically constraining them to learned mixtures of subspaces via efficient reparameterizations. We demonstrate scaling billion-parameter decentralized models to context lengths exceeding 100K tokens on networks as slow as 300Mbps, matching the wall-clock convergence speed of centralized models on 100Gbps interconnects.

AI 总结：针对联邦学习中混合专家模型因数据异质性导致的专家优化冲突问题，提出FC-MoE框架，通过重要性加权、梯度共识投影和局部知识保留机制，实现稳定优化并提升非独立同分布环境下的模型性能。

链接：https://arxiv.org/abs/2606.15625

机构：Nanyang Technological University（南洋理工大学）； University of Malaya（马来亚大学）

作者：Yijun Lu, Zihan Fang, Pengpeng Qiao, Zheng Lin, Jing Yang, Yuxin Zhang, Por Lip Yee, Zhe Chen, Jun Luo

英文摘要：The continuous scaling of large language models (LLMs) incurs prohibitive computational costs, making Mixture-of-Experts (MoE) a scalable alternative for efficient fine-tuning via sparse activation. While federated learning (FL) emerges as the paradigm for privacy-preserving collaborative optimization, integrating MoE into FL under data heterogeneity may trigger conflicting expert optimizations. Client-specific data distributions force same-indexed experts to optimize under inconsistent or even conflicting feature-label correlations. This mismatch induces destructive interference during aggregation, thus destabilizing the optimization trajectory and degrading model performance. To address this issue, we propose FC-MoE, a federated conflict-aware framework for MoE fine-tuning. It employs an importance aware weighting scheme to prioritize reliable local updates and utilizes gradient consensus projection to suppress conflicting updates, ensuring a stable global optimization path. Moreover, a local knowledge retention mechanism further preserves specialized client expertise by re-anchoring domain-specific residuals. Extensive experiments demonstrate that FC-MoE accelerates convergence and enhances both global and local model performance in non-IID federated environments.

AI 总结：针对异构联邦类增量学习中客户端标签子集不同、任务阶段不一致导致的旧知识遗忘问题，提出投影排练编排框架PRO及增强版PRO-MAX，通过服务器端维护紧凑类级投影记忆并实现平衡伪多任务训练，在图像、文本和图基准上提升异构流下的保留与最终效用。

链接：https://arxiv.org/abs/2606.15695

机构：VinUniversity； VNU-HCM, University of Science（胡志明市国家大学理科大学）； Technische Universität Berlin（柏林工业大学）

作者：Thinh T. H. Nguyen, Khoa D. Doan, Binh T. Nguyen, Danh Le-Phuoc, Kok-Seng Wong

英文摘要：Federated class-incremental learning (FCIL) becomes substantially harder when clients observe different label subsets, progress through tasks at different stages, and provide uneven supervision for the same semantic concepts. Existing FCIL methods often preserve old knowledge through input-space synthesis, but they can be fragile under heterogeneous task streams and difficult to transfer across modalities. To alleviate such issues, we propose PRO, a framework that replaces synthetic input replay with projected rehearsal orchestration. To remove external pretraining, we evaluate all methods under the same warmup. After this, PRO maintains compact class-level projected memories on the server and allows clients perform balanced pseudo multi-task training over current examples and old projected memories. To handle stronger representation drift, we further introduce PRO-MAX, which augments PRO with neighborhood-weighted memory alignment while preserving the same server-light principle that the server only aggregates model updates and memory statistics. Across image, text, and graph benchmarks, PRO and PRO-MAX improve retention and final utility under heterogeneous streams while remaining competitive in homogeneous FCIL. Even when baselines are given expanded replay budgets, they degrade under supervision imbalance and stage misalignment, indicating that replay quantity alone does not resolve replay-quality failures. Additional weak-task diagnostics further show that larger replay mismatch is associated with larger downstream degradation, while our method keeps projected memories better aligned with the evolving representation.

AI 总结：提出CaP-Eval工作流，在固定估计目标下审计合成学生数据的预测效用、因果保真度和隐私风险，发现DPGNet和蒸馏数据在保留处理效应结构上优于基线方法。

链接：https://arxiv.org/abs/2606.15940

作者：Hanghang Zheng, Xiwei Zhuang, Zhong Wang, Hong Liu, Xiao Chen, Jingwen He, Xia Li

英文摘要： Synthetic and distilled student data are increasingly used to enable privacy-conscious learning analytics, yet their suitability for decision-facing institutional support remains uncertain. In dropout support, generated data must preserve not only predictive utility or distributional resemblance, but also the financial-status evidence used to guide advising, payment-plan assistance, and scholarship-related decisions. Method: This study introduces CaP-Eval, a decision-facing causal-privacy audit workflow for evaluating generated student data under a fixed estimand, timing-aware adjustment design, estimator set, and empirical privacy-governance screen. The workflow compares original, distilled, adversarial synthetic, statistical synthetic, and DPGNet privacy-oriented generated data on predictive utility, treatment-effect fidelity, robustness to alternative estimators, and local training-record proximity. Results: DPGNet and distilled data preserved the original financial-status treatment-effect structure more reliably than the adversarial and Gaussian Copula baselines. DPGNet preserved full direction and rank agreement across epsilon levels; epsilon = 10 produced the smallest non-original IPW and DML deviations, while epsilon = 1 and epsilon = 5 amplified several financial-status contrasts. Distilled data remained highly faithful but retained the strongest local training-record proximity signal. TabularGNet preserved qualitative directions with moderate attenuation, and Gaussian Copula compressed effect magnitudes. Conclusions: Predictive utility, privacy orientation, empirical disclosure signals, and causal fidelity diverged; generated student data require joint audits of direction, magnitude, overlap, and release-governance risk before decision use.

AI 总结：针对隐私法规需求，提出首个实用通用审计框架，通过无知证明概念验证现有遗忘算法能否真正擦除指定数据影响，实验表明重训练和微调方法有效，去优化和Fisher/Hessian方法失败。

链接：https://arxiv.org/abs/2606.16110

机构：University of Technology Sydney（悉尼科技大学）； Deakin University（迪肯大学）； Macquarie University（麦考瑞大学）

作者：Dayong Ye, Tianqing Zhu, Ruiding Huang, Xinbo Fu, Jiayang Li, Bo Liu, Huan Huo, Wanlei Zhou

英文摘要：Machine unlearning has been extensively studied in response to growing privacy concerns and regulatory requirements. However, auditing whether unlearning algorithms have truly erased the influence of specific data remains an open challenge. The lack of reliable and practical auditing mechanisms can lead to critical privacy risks, such as residual information leakage. This paper initiates a systematic investigation into whether existing unlearning algorithms can truly forget the designated data. We propose the first practical and general-purpose auditing framework for machine unlearning, inspired by the concept of proof of ignorance. Our framework addresses the key practicality limitations of existing methods by eliminating the need for retraining-from-scratch baselines, avoiding the training of large numbers of shadow models, and requiring no intrusive intervention in the original training process. To evaluate the effectiveness of our framework, we first conduct validation experiments to verify its soundness and completeness. We then perform comprehensive experiments across six datasets and ten representative unlearning methods. The results demonstrate that our framework reliably distinguishes between successful and failed unlearning. In particular, we observe that retraining-based and fine-tuning-based methods can achieve effective unlearning, even when the target data remain in the original dataset. In contrast, de-optimization-based methods fail to achieve true unlearning and instead degrade the model's performance. Fisher/Hessian-based methods also fail to unlearn requested data, even formal certification is provided. Moreover, we show that our framework is robust against fake unlearning attempts and generalizes well to large language models.

AI 总结：揭示针对快速响应框架的投毒攻击，通过提示注入在训练集中植入恶意样本，实现目标性投毒和概念后门攻击，仅1%投毒率即可导致高达100%误报率和96%漏报率。

链接：https://arxiv.org/abs/2606.16242

机构：Princeton University（普林斯顿大学）

作者：David Huang, Jaewon Chang, Avidan Shah, Prateek Mittal, Chawin Sitawarin

英文摘要：The Rapid Response (RR) framework, deployed in production systems, including Anthropic's ASL-3 safeguards, continuously improves jailbreak-detection classifiers. When new jailbreaks emerge that bypass these classifiers, Rapid Response generates synthetic variants for training, helping the model generalize from the new attacks and quickly adapt. We reveal that prompt injection can infiltrate this pipeline to deliver poisoned samples into the classifier's training set, enabling two attack objectives: (I) targeted poisoning attacks that create false positives on harmless samples by categorizing them as a jailbreak, with a specific desired feature (e.g., certain formatting, subject, or keyword), (II) concept-based backdoor attacks that induce false negatives on jailbreak inputs, generalizing even to jailbreaks from attack strategies the defender explicitly trained against, when the backdoor trigger is present. Importantly, our threat model restricts adversaries to modifying only jailbreak samples (not benign data or labels), a constraint unexplored by prior work that makes the second objective particularly challenging. We address this with Omission Attack, which exploits a new phenomenon: when training on concept-absent unsafe samples, the classifier misassociates that concept's presence with the safe label. Both attacks cause substantial and in some cases near-complete label flipping at only a 1% poisoning rate, achieving up to 100% false positive rates and up to 96% false negative rates.

AI 总结：针对个性化联邦学习（pFL）中共享层与个性化层分离带来的遗忘挑战，提出pFedUL框架，通过梯度贡献归因、自适应选择性遗忘和轻量级重校准协议，在保证遗忘效果的同时维持剩余客户97.3%的个性化精度。

链接：https://arxiv.org/abs/2606.16304

作者：Zhuodong Liu, Xiangyu Li, Zhihao Zhang

英文摘要： Federated unlearning (FU) enables the removal of specific data contributions from federated learning (FL) models to comply with regulations such as the General Data Protection Regulation (GDPR). However, most existing FU methods are designed for the FedAvg paradigm, where all clients share a single global model. In practice, personalized federated learning (pFL) methods such as FedPer, FedRep, Ditto, and FedBN have become widely adopted due to their superior handling of non-IID data. These methods decompose the model into shared global layers and client-specific personalized layers, fundamentally altering the semantics of unlearning, yet this setting has received little attention. We formalize FU under the pFL paradigm, identifying a tension between unlearning completeness on shared layers and personalization preservation for remaining clients. We then propose pFedUL, a layer-aware selective unlearning framework comprising three components: (1) gradient-based layer-wise contribution attribution that separately quantifies the target client's influence on shared and personalized parameters, (2) adaptive selective unlearning that applies differentiated forgetting strategies across layer types, and (3) a lightweight recalibration protocol enabling remaining clients to restore personalization with minimal overhead. We further introduce two new metrics, Personalization Preservation Score (PPS) and Cross-client Fairness Index (CFI), to evaluate pFL-specific unlearning quality. Experiments on CIFAR-10, CIFAR-100, and FEMNIST under varying non-IID settings indicate that pFedUL achieves unlearning effectiveness comparable to full retraining while maintaining an average of 97.3\% personalized accuracy for remaining clients. Compared with six state-of-the-art FU methods adapted to the pFL setting, pFedUL consistently achieves superior personalization preservation.

AI 总结：针对拆分推理中隐藏表示易被近邻搜索恢复的问题，提出正交混淆方法，并设计ConjFormer架构实现O(d)-等变性，在不加噪声或重加密下将令牌恢复率从35%降至1.3%，困惑度仅增0.4%。

链接：https://arxiv.org/abs/2606.16461

机构：MBZUAI（穆罕默德·本·扎耶德人工智能大学）； University of Southern California（南加州大学）

作者：Alexander Yukhimchuk, Andrey Shulga, Mladen Kolar, Martin Takáč

英文摘要：Running large language models locally is often impractical, pushing inference on sensitive text to third-party providers. Split inference partially mitigates this by keeping tokens on the client and sending only hidden representations, but these representations can still be recovered via nearest-neighbor search against the public embedding table. We propose an orthogonal obfuscation procedure in which the client multiplies embeddings by a secret orthogonal matrix before transmission. To enable correct inference under arbitrary rotations, we introduce ConjFormer, a transformer variant that is exactly $\mathrm{O}(d)$-equivariant via a lightweight normalization change (scalar RMSNorm) together with blockwise orthogonal conjugation of all linear weights. As a result, the server performs the full forward pass entirely in the rotated basis and never observes unrotated hidden states. Experiments on GPT-2 and Llama 3.2 1B models fine-tuned on PubMed show that orthogonal obfuscation eliminates direct cosine nearest-neighbor inversion and reduces token recovery from over 35% top-10 to at most 1.3%, while increasing perplexity by only 0.4% after fine-tuning. These results indicate that enforcing symmetry at the architectural level can provide a practical defense for privacy-preserving LLM inference without noise injection or heavy cryptographic machinery.

AI 总结：针对单轮联邦学习中客户端数据异构导致的特征错位问题，提出SLOT-Align方法，利用共享冻结编码器、Bures-Wasserstein重心和测地最优传输映射实现无训练的特征对齐，提升模型精度与鲁棒性。

链接：https://arxiv.org/abs/2606.16655

机构：AI for Good (AIGO), Italian Institute of Technology（意大利技术研究院AI for Good (AIGO)）； MaLGa-DIBRIS, University of Genoa（热那亚大学MaLGa-DIBRIS）； Department of Computer Science, University of Verona（维罗纳大学计算机科学系）

作者：Daniele Berardini (1), Vito Paolo Pastore (1 and 2), Vittorio Murino (1 and 3) ((1) AI for Good (AIGO), Italian Institute of Technology, Genoa, Italy, (2) MaLGa-DIBRIS, University of Genoa, Genoa, Italy, (3) Department of Computer Science, University of Verona, Verona, Italy)

英文摘要：One-Shot Federated Learning (OSFL) addresses extreme communication regimes in which clients interact with the server only once, amplifying the impact of heterogeneous client data distributions. In particular, the interaction of domain shift and label shift across clients induces misaligned feature representations that cannot be corrected through iterative optimization. Existing OSFL methods rely on distillation, server-side generation or ensemble-based aggregation, but assume aligned representations or address domain and label shift separately. We introduce SLOT-Align (Single-round, Learning-free Optimal Transport Alignment), a geometry-aware feature harmonization framework for OSFL. SLOT-Align uses a shared frozen encoder to extract compact feature statistics, constructs a global reference via Bures-Wasserstein barycenters, and aligns local representations using closed-form geodesic optimal transport maps. The method is computationally efficient and can be combined with existing OSFL pipelines relying on frozen encoders without modifying their training procedures. Extensive experiments across multiple benchmarks, pretrained backbones, and OSFL methods show that SLOT-Align consistently improves accuracy and robustness under joint domain and label shift.

AI 总结：本文提出联邦消息的正式数学定义，建立包含模型结构、统计摘要和数据条件表示的三类分类法，分析计算、通信和隐私权衡，并综述202篇文献揭示2021年后消息范式多样化趋势。

链接：https://arxiv.org/abs/2606.16891

机构：AIMS lab, Center for Neurosciences, UZ Brussel, Vrije Universiteit Brussel, Brussels, Belgium（AIMS实验室，神经科学中心，布鲁塞尔大学医院，布鲁塞尔自由大学，布鲁塞尔，比利时）； Artificial Intelligence Lab, Vrije Universiteit Brussel, Brussels, Belgium（人工智能实验室，布鲁塞尔自由大学，布鲁塞尔，比利时）

作者：Alvaro Javier Vargas Guerrero, Xinguang Wang, Quang Manh Doan, Guy Nagels

英文摘要： Federated Learning is rapidly evolving beyond the exchange of traditional model weights and gradients, yet existing definitions fail to capture the full scope of modern payloads like synthetic data and federated analytics. This paper addresses the gap by proposing a formal mathematical definition of a federated message that accounts for both utility and privacy. We introduce a taxonomy that organizes these exchanges into three categories: model structures, statistical summaries, and data-conditioned representations. By evaluating these groups based on computational demands, communication costs, and privacy risks, we provide a clearer understanding of the trade-offs involved in decentralized training. Our review of 202 recent publications highlights a significant shift since 2021 toward diverse messaging paradigms, signaling a move away from standard deep learning updates toward more specialized information sharing. This framework provides a structured path for future research to optimize federated systems for varying hardware and security requirements.

AI 总结：提出一个可定制的实证审计框架，通过区分真实披露与幻象披露，利用统计假设检验检测合成数据中的隐私泄露，无需模型访问或参考模型，提供比先前方法更紧的隐私泄露下界。

链接：https://arxiv.org/abs/2606.16952

机构：Google（谷歌）； University of Southern California（南加州大学）

作者：Kareem Amin, Rudrajit Das, Alessandro Epasto, Adel Javanmard, Dennis Kraft, Mónica Ribero, Sergei Vassilvitskii

英文摘要：The rapid adoption of generative AI and Large Language Models (LLMs) has spurred interest in synthetic data as a privacy-preserving alternative to sensitive real-world datasets. However, generating high-utility synthetic data often carries the risk of memorizing and regurgitating private information from the training corpus. In this work, we present a customizable empirical auditing framework designed to detect and explain such data disclosures. Our framework introduces a mechanism to distinguish between "true disclosures"-where the system directly reproduces a user's information-and "phantom disclosures''-where the system incidentally generates a user's data. By partitioning input data into training and holdout sets and applying rigorous statistical hypothesis testing, we determine if observed disclosures are consistent with strict privacy baselines, such as zero-learning or specific Differential Privacy (DP) bounds. Crucially, this approach requires no model access, no canary insertion, and no reference model training -only the synthetic output and a held-out control set. We demonstrate that this framework effectively functions as a membership inference attack, providing empirical lower bounds on privacy leakage that are tighter than prior data-based auditing methods. Our approach is model-agnostic, applies to any synthetic data generation mechanism, and requires orders of magnitude fewer computational resources than shadow-model or canary-based alternatives.

AI 总结：针对差分隐私联邦学习，提出RING攻击，利用差分隐私的掩蔽效应绕过防御，在中等隐私预算下平均攻击成功率90.3%。

链接：https://arxiv.org/abs/2606.17035

作者：Xiaolin Li, Ning Wang, Ninghui Li, Wenhai Sun

英文摘要：Prior research suggests that differential privacy (DP) inherently enhances the robustness of federated learning (FL) against backdoor attacks. In this paper, we challenge this assumption. Through an empirical analysis of two baseline attack strategies, we uncover a fundamental tension in DP-FL: while bypassing DP allows state-of-the-art defenses to detect and filter malicious updates, complying with DP inadvertently masks their distinguishing statistical characteristics. Consequently, existing defenses become ineffective as DP reduces the raw backdoor signal. Building on this masking effect, we propose RING, a novel attack that explicitly exploits DP to conceal malicious contributions while maximizing attack impact. By collaboratively crafting adversarial perturbations, compromised clients reconstruct a strong backdoor signal during aggregation without triggering anomaly detection. RING operates as a perturbation layer that is agnostic to the underlying backdoor technique, making it broadly applicable and composable with existing attacks -- a property that significantly amplifies the threat it poses to DP-FL. Extensive evaluations across four image and text datasets under non-iid distributions show that RING achieves an average attack success rate of 90.3% against six state-of-the-art defenses under a moderate privacy budget, an improvement of up to 26.08x over baseline strategies. Finally, we evaluate potential countermeasures and find that mitigating this threat incurs significant utility trade-offs, exposing a fundamental security gap in the deployment of differentially private FL.

AI 总结：提出GRAPE框架，通过逐步暴露参数空间并利用对抗谱利用分数引导容量分配，在固定计算预算下提升紧凑模型的对抗鲁棒性，在CIFAR-10上以1.009倍FLOPs将PGD-20鲁棒准确率从51.70%提升至56.94%，参数减少21.4%。

链接：https://arxiv.org/abs/2606.14865

机构：University of Science and Technology of China（中国科学技术大学）； China Mobile (Suzhou) Software Technology Co., Ltd.（中移（苏州）软件技术有限公司）

作者：Zhiyuan Ye (1), Xiangyu Zhou (2), Ji Qi (2), Hao Zhang (1), Yi Zhou (2) ((1) University of Science and Technology of China, (2) China Mobile (Suzhou) Software Technology Co., Ltd.)

英文摘要： Adversarial Training (AT) improves neural network robustness, but most methods train a fixed parameter space from the start. This paper asks whether the order in which parameters become optimizable can affect the final robust solution, even when the final architecture or computation budget is controlled. We propose GRAPE, Guided Parameter-Space Evolution, a training framework for compact adversarial robustness. GRAPE combines parameter-space stabilization with progressive hidden expansion: it stabilizes robust optimization in the currently exposed space, gradually releases new optimizable dimensions, and uses an adversarial spectral utilization score to guide newly released capacity toward high-pressure modules. In contrast to fixed-structure AT, GRAPE treats robust model learning as a process of progressive parameter-space exposure and evolution. Under the standard $\ell_\infty$ threat model on CIFAR-10, with fixed-structure ResNet-18 AT as a controlled reference, GRAPE improves PGD-20 robust accuracy from 51.70% to 56.94% at a nearly matched computation budget with a FLOPs ratio of 1.009x, while reducing parameter count by about 21.4%. A sequential grow variant with the same final ResNet-18 architecture reaches 56.52% PGD-20 robust accuracy, indicating that the gain is not only due to final architecture differences but also to the parameter-space exposure path. These results suggest that guided parameter-space evolution can yield compact and robust parameter configurations under matched computation.

AI 总结：针对仅用准确率评估忽略推理链中偏见承认的问题，提出包含易感性（susceptibility）和承认（acknowledgment）两个维度的诊断方法，实验发现不同模型在准确率相近时承认率差异显著。

链接：https://arxiv.org/abs/2606.15127

作者：Xian Sun, Wei Gao, Yingshuo Wang, Lingdong Kong, Yanhang Li, Zhichao Fan, Zexin Zhuang, Wenlong Dong, Zhiyuan Zheng, Hrishikesh Paranjape, Abhishek Mandal, Johnny R. Zhang

英文摘要：Reasoning models are increasingly used in settings where the final answer is not the only object of review: educational tools may show students intermediate steps, decision-support systems may require human oversight, and audit workflows may inspect traces for misleading or biased input. In such settings, two responses can receive the same final-answer score while differing in whether the trace explicitly flags injected biasing content. Accuracy-only evaluation collapses these cases. We study this gap as a measurement blind spot for responsible evaluation and introduce a minimal trace-level diagnostic with two axes: \emph{susceptibility} (whether the bias breaks a previously correct answer) and \emph{acknowledgment} (whether the trace contains a rubric-defined surface reference to the injected content). Across thousands of biased GSM8K trials, GPT-4o and Claude Sonnet~4 have similar susceptibility rates ($1.3\%$ vs.\ $1.2\%$) but substantially different acknowledgment rates ($13.0\%$ vs.\ $75.0\%$) under the same rubric.

AI 总结：审计分布自由风险控制下选择性预测的边界紧致性与可交换性假设，发现经验阈值法常超预算，而认证方法在可交换时有效，但组部署下因可交换性失效导致违规。

链接：https://arxiv.org/abs/2606.15153

作者：Jingwen Zhou, Mingzhe Wang

英文摘要：Selective prediction with distribution-free risk control promises that, with confidence 1-delta over the calibration draw, the error rate of accepted inputs stays below a user budget alpha. We audit this promise on signal-domain detectors -- machine anomalous-sound detection (ASD) and AI-generated-image forensics -- for four calibration rules: uncertified empirical thresholding (NAIVE) and certified Hoeffding, Clopper-Pearson (CP), and betting (WSR) upper confidence bounds. We report three findings. (i) NAIVE thresholding, common in practice, exceeds its declared budget in 49-73% of synthetic trials (n=200 calibration points) and in up to 68% of real-data splits: a false sense of safety rather than a broken theorem, since the rule never had a certificate. (ii) Tightness matters: CP and WSR certify substantial coverage where Hoeffding certifies none, with zero observed budget overruns under exchangeable splits. (iii) Under grouped deployment (unseen machine types or generators), certified rules overrun in 9-30% of trials -- far above delta -- showing the failure lies in the broken exchangeability premise, not in the bounds; a conservative per-group threshold restores validity at a severe coverage cost.

AI 总结：提出贝叶斯可转向CNN，通过后验分布赋予核随机性同时保持SE(3)-等变性，实现不确定性分解，在分类精度和分布偏移下鲁棒性优于确定性模型。

链接：https://arxiv.org/abs/2606.15479

机构：Michigan Technological University（密歇根理工大学）； Johns Hopkins University（约翰霍普金斯大学）； The Center for Artificial Intelligence at the Institute of Computing and Cybersystems, Michigan Technological University（密歇根理工大学计算与网络系统研究所人工智能中心）

作者：Abhishek Keripale, Ponkrshnan Thiagarajan, Susanta Ghosh

英文摘要：Steerable convolutional neural networks (Steerable-CNNs) guarantee SE(3)-equivariance by parameterizing kernels as linear combinations of steerable basis functions, but their deterministic nature precludes uncertainty quantification - limiting their use in settings where confidence estimates are essential. We propose a Bayesian Steerable-CNN that places posterior distributions over the basis coefficients, yielding stochastic kernels while preserving equivariance exactly. The loss function of the model is obtained via variational inference and minimized by Bayes-by-Backpropagation. The framework admits a decomposition of predictive uncertainty into epistemic and aleatoric components. Empirically, the model attains competitive classification accuracy alongside an expected calibration error of 0.0263 and outperforms its deterministic counterpart by up to 6.17% under distributional shift induced by additive Gaussian noise. Furthermore, we leverage the model's uncertainty estimates to enhance its performance significantly, achieving a notable gain - approximately 4% higher accuracy across 84% of the test dataset. A statistically significant negative correlation between epistemic uncertainty and prediction error confirms that the learned posterior variance is semantically meaningful. The framework unifies Bayesian uncertainty quantification with the inductive bias of equivariant CNNs.

AI 总结：本文通过计算替代模型的Rashomon集并评估其多样性，发现高保真替代模型在关键性能指标上可能与目标模型存在显著差异，挑战了传统观点。

链接：https://arxiv.org/abs/2606.15493

机构：ÉTS, Mila（蒙特利尔高等技术学院，Mila）； The University of Electro-Communications（电气通信大学）

作者：Eliott Baltz, Satoshi Hara, Ulrich Aïvodji

英文摘要： Model stealing attacks, where adversaries create high-fidelity surrogate models, are a significant threat to the intellectual property of machine learning services. Conventional wisdom suggests these surrogates could provide adversaries with economic leverage comparable to the original service providers. This paper challenges this assumption by evaluating model stealing attacks beyond mere fidelity to the target model. Because query-based extraction provides only partial supervision of the target's input-output behavior, the surrogate is not uniquely identified: many near-optimal surrogates can achieve comparable fidelity while differing in deployment-relevant properties. Instead of performing a classic learning-based model stealing attack, we compute the Rashomon Set (i.e., the set of almost-equally-accurate models) of surrogate models, and evaluate its diversity using multiplicity metrics (ambiguity, discrepancy, and Rashomon Capacity) and group fairness metrics. Across tabular, medical imaging, and NLP tasks, our experiments on real-world datasets reveal that despite exhibiting similar fidelity to the target model, surrogate models can display significant variances in other critical performance metrics. These findings cast doubt on the presumed equivalence between high-fidelity surrogates and the target model in practical deployment scenarios.

AI 总结：提出贪婪坐标扩散方法，利用扩散模型引导生成语义连贯的对抗样本，在保持自然性的同时实现高攻击成功率。

链接：https://arxiv.org/abs/2606.15531

作者：Bohdan Turbal, Blossom Metevier, Max Springer, Aleksandra Korolova

英文摘要：Fine-tuning aligned language models on benign tasks (e.g. math tutoring) systematically breaks safety guardrails, even when training data contains no harmful content. While mechanistic approaches have shed light on where alignment resides in model weights, they do not by provide a general formal framework for deriving guarantees about when fine-tuning degrades it -- leaving the field without principled tools for predicting or preventing alignment collapse. We develop a local geometric framework through geometric analysis of parameter-space trajectories and apply it to understand the fragility of alignment in fine-tuning. While first-order analysis suggests orthogonal updates are safe, we prove this is illusory: the curvature of the fine-tuning loss induces second-order acceleration that can induce second-order drift into alignment-sensitive regions. We formalize a construct of our framework as the Alignment Instability Condition (AIC), three geometric properties that, when present, are sufficient to guarantee degradation. Our main result proves quartic onset of alignment degradation along gradient-flow trajectories, determined by how sharply alignment depends on specific parameters and how strongly tasks couple to these parameters. These findings yield formal sufficient conditions under which static first-order protection can fail under gradient descent. We further empirically validate the framework's foundations, showing that the Fisher Information Matrix provides a proxy for the degree of safety degradation across diverse fine-tuning.

AI 总结：提出InstantForget方法，通过推理时特征重置实现无需参数更新的后门遗忘，利用马氏距离检测异常特征并重置为中性表示，在CIFAR-10上平均ASR降至0.071。

链接：https://arxiv.org/abs/2606.15730

机构：College of Computer Science and Artificial Intelligence, Fudan University（复旦大学计算机科学与人工智能学院）

作者：Zhenyu Yu

英文摘要：Backdoor unlearning aims to remove a malicious trigger behavior from a deployed model while preserving clean utility. We study the update-free inference-time setting, where model parameters remain frozen. First, we audit a common projection assumption under oracle paired clean and triggered features. Projection succeeds mainly on BadNets and leaves WaNet, Blended, and SIG at 0.683, 0.888, and 0.941 ASR on CIFAR-10 ResNet-18. This failure is not explained by spectral compactness, spatial locality, or subspace misalignment. It is predicted by a logit-triplet gap involving the target margin, target-logit drop, and non-target logit rise. We then introduce InstantForget, a clean-calibrated gated reset that flags anomalous features with a Mahalanobis score and moves only flagged features toward a neutral non-target representation. With one fixed operating point selected on held-out triggered validation, InstantForget reduces average ASR to 0.071 across four non-adaptive CIFAR-10 triggers without triggered samples or parameter updates at deployment. It also reaches 0.981 detection AUROC and transfers to six of eight tested backbones. Reported failures under WaNet, ModelNet10 point blend, two backbone geometries, and adaptive feature-compactness attacks define the method's scope.

AI 总结：提出不确定性激活图（UAM）框架，结合证据深度学习与全梯度类激活映射，生成空间不确定性激活图，区分缺乏证据的空虚和假设冲突的不和谐，填补不确定性量化与可解释性之间的空白。

链接：https://arxiv.org/abs/2606.15767

机构：University of the District of Columbia（哥伦比亚特区大学）； University of Texas at Dallas（德克萨斯大学达拉斯分校）； Virginia Tech（弗吉尼亚理工大学）； U.S. Army DEVCOM Army Research Laboratory（美国陆军DEVCOM陆军研究实验室）； University of Oslo（奥斯陆大学）； Bowie State University（鲍伊州立大学）

作者：Dong Hyun Jeong, Feng Chen, Jin-Hee Cho, Lance M. Kaplan, Audun Jøsang, Soo-Yeon Ji

英文摘要： Understanding when and why deep neural networks are uncertain is crucial for deploying reliable machine learning systems in safety-critical domains. While existing uncertainty quantification methods provide scalar measures of model confidence, they offer limited insight into which spatial regions of an input contribute to different types of uncertainty. We propose a novel visualization framework, Uncertainty Activation Map (UAM), that combines Evidential Deep Learning (EDL) with Full-Gradient Class Activation Mapping (FullGrad) to generate interpretable spatial uncertainty activation maps. Our approach distinguishes between two fundamental types of uncertainty: vacuity, representing lack of evidence, and dissonance, capturing conflicting evidence between competing hypotheses. By leveraging the complete gradient decomposition property of FullGrad and the principled uncertainty quantification of Subjective Logic, our method produces theoretically grounded visualizations that highlight specific image regions responsible for model uncertainty. With this framework, vacuity and dissonance activation maps are generated by computing belief-weighted attributions, enabling identification of where models lack knowledge versus where they encounter ambiguous evidence. Extensive evaluations across multiple benchmark datasets demonstrate that the proposed framework effectively addresses the critical gap between uncertainty quantification and explainability, providing intuitive visual feedback to assess model reliability in complex visual recognition tasks.

AI 总结：研究语言模型更新后激活监控器是否仍可靠，发现量化更新影响小，微调更新常导致监控器失效，且可通过预部署特征预测退化。

链接：https://arxiv.org/abs/2606.15980

机构：University of Michigan（密歇根大学）

作者：Evan Duan

英文摘要：Activation monitors-lightweight probes trained on a language model's internal representations-are an increasingly common layer in deployment safety stacks. Deployed models however are rarely static: they are quantized, fine-tuned, adapted with LoRA, or served with merged adapters while the monitor remains frozen. We present the first systematic test of whether this implicit contract holds: whether activation monitors trained on a base model remain reliable after these routine model updates. Across multiple safety-relevant monitors, model depths, update families, and open-weight models, we find a sharp split: quantization-style updates largely preserve frozen probe performance, while fine-tuning-style updates frequently make probes stale. Fragility is highly monitor-dependent, with privacy/PII probes most affected and refusal-compliance probes comparatively stable, showing that retraining a behavior need not stale its corresponding monitor. QLoRA is especially damaging despite NF4 quantization alone being relatively benign, suggesting that quantization becomes riskier when combined with adaptation. We further show that degradation is predictable from pre-deployment features, enabling revalidation budgets to be triaged toward the monitors most likely to fail. These results suggest that fine-tuning should trigger activation-monitor revalidation by default, while prediction can help prioritize which monitors to check first.

AI 总结：提出自适应对数相关熵损失（ALCL），通过可微重参数化联合学习形状和尺度参数，使损失几何动态适应残差统计，抑制极端异常值，在混合重尾和脉冲噪声下优于MSE和固定核相关熵损失。

链接：https://arxiv.org/abs/2606.16050

机构：University of South Florida（南佛罗里达大学）； California Polytechnic State University（加州州立理工大学）

作者：Mainak Kundu, Ria Kanjilal, Ismail Uysal

英文摘要：Robust deep learning under heavy-tailed and impulsive noise remains challenging because conventional losses such as mean squared error (MSE) exhibit unbounded sensitivity to outliers. Although correntropy-based objectives improve robustness, existing formulations rely on fixed kernel parameters that must be empirically tuned and remain static during training. To address these limitations, we propose an Adaptive Log-Correntropy Loss (ALCL), a heavy-tailed loss formulation that adaptively learns its robustness geometry during optimization. ALCL introduces a logarithmic residual model whose shape and scale parameters are learned jointly with network weights through differentiable reparameterization. This yields a principled maximum likelihood formulation whose influence function is formally bounded and redescending, allowing the loss geometry to adapt dynamically to evolving residual statistics while suppressing extreme outliers. Comparative experiments on four widely used benchmark datasets spanning grayscale and red-green-blue (RGB) image data under mixed heavy-tailed and impulsive noise demonstrate that ALCL consistently outperforms MSE and optimally tuned generalized correntropy losses in both reconstruction fidelity and downstream classification accuracy. While performance differences remain small under low-noise conditions, under high-noise regimes ALCL improves median accuracy by up to 4.75% on grayscale benchmarks and 4.51% on RGB datasets, with reduced variance across runs. These results demonstrate that adaptive robustness through joint learning of loss parameters provides a computationally efficient alternative to static correntropy-based losses for deep learning in non-Gaussian environments.

AI 总结：提出一种基于类条件语义扰动和稀疏自编码器的可解释OOD检测框架，通过分析表示稳定性实现检测与内部机制解释。

链接：https://arxiv.org/abs/2606.16196

机构：NepAl Applied Mathematics and Informatics Institute for research（尼泊尔应用数学与信息学研究所）； West Virginia University（西弗吉尼亚大学）； Kathmandu University（加德满都大学）； University College London（伦敦大学学院）； University of Aberdeen（阿伯丁大学）

作者：Anju Chhetri, Pratik Shrestha, Ramesh Rana, Prashnna Gyawali, Binod Bhattarai

英文摘要： Deep neural networks have achieved remarkable performance across medical imaging tasks, yet their tendency to overgeneralize under distributional shifts poses a major obstacle to safe clinical deployment. Out-of-Distribution (OOD) detection methods aim to mitigate this risk, but most existing approaches rely on opaque internal signals with poorly understood semantic meaning, limiting trust in safety-critical settings. In this work, we propose an interpretable OOD detection framework that probes the stability of model predictions under class-conditioned semantic perturbations. Leveraging sparse autoencoders (SAEs), we learn class-specific concept vectors from in-distribution data that disentangle dense intermediate representations into sparse, semantically meaningful components. At inference, we perturb deeper-layer representations using the concept vectors associated with the model's predicted class and measure the class logits stability. We hypothesize that in-distribution samples exhibit low sensitivity to such perturbations, as their representations align with class-specific semantic directions, whereas OOD samples show amplified deviations due to representational misalignment. By framing OOD detection as a concept conditioned stability analysis, our approach provides both a discriminative OOD signal and an interpretable lens into the internal mechanisms driving model uncertainty, making it particularly suitable for high stakes medical applications.

AI 总结：提出校准方差传播（CVP），通过新型归一化层传播方法、激活函数处理技术及轻量校准步骤，在单次前向传播中高效估计不确定性，在Transformer和CNN上达到与MC采样相当的精度，成本显著降低。

链接：https://arxiv.org/abs/2606.16214

机构：TU Darmstadt & hessian.AI, Darmstadt, Germany（达姆施塔特工业大学 & hessian.AI，德国达姆施塔特）； RIKEN Center for Advanced Intelligence Project, Tokyo, Japan（日本理化学研究所革新智能研究中心，日本东京）

作者：Tobias Jan Wieczorek, Leon de Andrade, Thomas Möllenhoff, Marcus Rohrbach

英文摘要：Modern deep learning models remain notoriously prone to overconfidence, limiting their reliability in high-stakes applications. Bayesian methods aim to counter this by learning a distribution over model parameters, and recent advances now make this feasible for large-scale architectures at costs comparable to AdamW. However, a challenge remains at test time: predictions must be averaged across many forward passes with weights sampled from the posterior, which is prohibitively expensive. Variance propagation offers an efficient alternative, computing layer-wise analytical approximations of uncertainty in a single forward pass. While such techniques are effective for MLPs, their extension to modern architectures remains challenging, due to increased depth and diversity of layer types. To fill this gap, we propose Calibrated Variance Propagation (CVP), which introduces a new propagation method for normalization layers, combines it with recent techniques for handling activation functions, and absorbs residual error through a light calibration step. CVP yields comparably accurate uncertainty estimates to MC sampling across transformers and CNNs, at a fraction of the cost. Against prior variance propagation work, CVP improves coverage at $0.5\%$ risk from $8.2\%$ to $14.6\%$ with BEiT-3 on Visual Reasoning (NLVR2) and from $2.6\%$ to $10.8\%$ with ViLT on VQAv2, with gains extending to convolutional architectures.

AI 总结：提出神经贝叶斯异常缓解（NBAM）损失，基于贝叶斯潜变量混合模型，既提供鲁棒监督损失又输出无监督污染后验，在CIFAR-10上优于Huber等基线。

链接：https://arxiv.org/abs/2606.16524

机构：Astrophysics Group, Cavendish Laboratory, University of Cambridge（剑桥大学卡文迪许实验室天体物理组）； Institute of Astronomy, University of Cambridge（剑桥大学天文研究所）

作者：S. A. K. Leeney, W. J. Handley, H. T. J. Bevins, E. de Lera Acedo

英文摘要：Engineered robust losses such as Huber, Student-$t$, and generalised cross-entropy make supervised models tolerant of contamination but cannot answer which observations are corrupted. We introduce Neural Bayesian Anomaly Mitigation (NBAM), a general-purpose drop-in loss derived from a Bayesian latent-switch mixture model: the marginal likelihood defines a robust supervised loss, and the associated posterior defines an unsupervised contamination classifier. Like Huber or Student-$t$, NBAM can replace the standard training loss in any supervised pipeline; unlike them, it additionally learns a structured contamination model and returns a calibrated per-sample contamination posterior. A learned input-dependent prior $\pi_\phi(x)$ captures the spatial locality of contamination, so that samples near known corruptions are more likely to be flagged, while an Occam penalty emerges automatically and regularises against over-flagging. On CIFAR-10 with asymmetric label contamination, NBAM recovers the structure of the corruption process without supervision: the contamination posterior separates clean from corrupted samples, and the learned anomaly head identifies the direction of every label-flip pair. Alongside these capabilities, NBAM outperforms the four robust-loss baselines considered here at contamination rates 0.2-0.6.

AI 总结：本文研究概念瓶颈模型（CBM）中符号检测的可靠性问题，通过交换独立训练的概念检测器和分类头来识别易受虚假激活影响的概念，并提出一种可靠性感知训练策略，在CUB-200-2011和合成任务上验证了其有效性。

链接：https://arxiv.org/abs/2606.16535

机构：University of Essex（埃塞克斯大学）

作者：Javier Fumanal-Idocin, Javier Andreu-Perez

英文摘要：Concept Bottleneck Models (CBMs) are a relevant tool for explainable Artificial Intelligence because they make their predictions through human-interpretable symbols. However, high task accuracy does not guarantee that these symbols are detected faithfully: jointly trained CBMs may encode task-specific shortcuts in the bottleneck, making their explanations unreliable. In this paper, we study concept-detection reliability by swapping independently trained concept detectors and classification heads that share the same symbolic vocabulary. We use the resulting performance degradation, concept-level metrics, and symbol-wise uncertainty estimates to identify concepts that are especially prone to spurious firing. Finally, we propose a reliability-aware training strategy in which a shared concept detector is optimized with multiple classification heads and penalized for relying on globally or instance-wise unreliable symbols. On CUB-200-2011 with full concept supervision, detectors and heads are almost freely interchangeable (swap drop below one accuracy point, relative retention above $99\%$, and no concept detected below chance), whereas on a controlled synthetic task we show that, as the concept-supervision weight is reduced, models keep near-perfect task accuracy while swapped accuracy and agreement with the ground-truth concepts collapse to chance. Our reliability-aware training substantially mitigates this leakage, roughly doubling swap accuracy in the leaky regime.

AI 总结：针对神经算子模型从仿真到现实迁移时的精度下降问题，提出PhysGuard框架，利用仿真数据的Fisher信息矩阵保护物理关键参数，限制微调更新方向，在严重域偏移下将低频误差降低32%。

链接：https://arxiv.org/abs/2606.16602

机构：Faculty of Engineering and IT, University of Melbourne（墨尔本大学工程与信息技术学院）； School of Computing, National University of Singapore（新加坡国立大学计算机学院）； Artificial Intelligence Research Institute, IFLYTEK Co., Ltd.（科大讯飞股份有限公司人工智能研究院）

作者：Changjian Zhou, Junfeng Fang, Negin Yousefpour, Peng Wu, Bin Yan, Guillermo A Narsilio

英文摘要：Neural operator models trained on simulation data often lose accuracy when applied to experimental measurements due to the sim-to-real gap. Standard fine-tuning with limited real data can reduce this gap, but it may also damage the core physics-relevant representations learned during pretraining. Although knowledge-preserving adaptation has been widely investigated in vision or language tasks, it remains unclear whether these methods are suitable for neural operators whose architectures and protected knowledge are fundamentally different. Neural operators need to preserve core-scale physical structures rather than semantic or visual features. We propose PhysGuard, a physics-preserving framework for accurate sim-to-real adaptation of neural operators. Specifically, PhysGuard uses the empirical Fisher Information Matrix computed on simulation data to identify physics-critical parameter directions, then restricts fine-tuning updates to directions that do not interfere with them. A layer-wise Gram-matrix formulation makes this efficient for models with millions of parameters, while an adaptive threshold automatically determines the protected subspace size. A spectral probe experiment shows that the dominant Fisher directions are strongly associated with low-frequency output structures. Experiments on benchmark across four neural operator architectures and different physical systems show that PhysGuard performs strongly on most evaluation metrics compared to baselines. The benefits are most evident under severe domain shift, where it reduces low-frequency error by up to 32\% compared to standard fine-tuning while maintaining adaptability. Our code is available at this https URL.

AI 总结：研究多模态自评估中偏好坍缩的加剧现象，发现跨模态传染导致策略选择扭曲，并引入传染矩阵量化风险。

链接：https://arxiv.org/abs/2606.16682

机构：Qilu Institute of Technology, School of Software Engineering（齐鲁理工学院软件工程学院）

作者：Zewen Liu

英文摘要：When AI agents use language models to evaluate their own outputs in a feedback loop, systematic biases emerge. We show that Evaluator Preference Collapse (EPC) is dramatically amplified in multimodal settings. Using GPT-4o to evaluate DeepSeek-chat across text and visual tasks, we find that a single strategy (step_by_step) absorbs 48.4% of all weight -- 3.2x the collapse observed in text-only self-evaluation -- while three visual-domain strategies receive only 9.1% combined weight. We then demonstrate a novel phenomenon we term cross-modal contagion: evaluator preferences acquired on one modality transfer to and corrupt strategy selection on another. Through a four-phase isolation training paradigm, we measure contagion coefficients and document strategy inversion -- the optimal strategy for a modality reverses after cross-modal exposure. A Phase 3 statistical validation across four evaluator configurations (N=53 total independent repetitions, 15,592 API calls) reveals a clear hierarchy: cross-model evaluation (GPT-4o, N=8) produces strong but symmetric bidirectional contagion (mean gamma_{T->V}=1.176, gamma_{V->T}=1.089, Delta=-0.088, p=0.575, Cohen's d=0.29); high round counts (DashScope, 50 rounds) cause collapse to single-strategy dominance (70% zero contagion); and self-evaluation provides near-complete immunity -- 97% of runs (N=30, DeepSeek-chat) yield exactly zero contagion (mean gamma=0.033, 95% CI [-0.031, 0.010], p=0.642, d=0.07). No evaluator condition shows statistically significant directional asymmetry. We introduce the contagion matrix indexed by evaluator identity, release the MM-EPC experimental framework, and identify cross-model evaluator architecture as the primary risk factor for preference contagion.

AI 总结：针对算法解释在实践中含义模糊且信息不足的问题，提出解释卡，通过补充鲁棒性和有效性信息及解释说明，帮助用户正确解读，并满足欧盟AI法案的可解释性要求。

链接：https://arxiv.org/abs/2606.16786

机构：University of Tübingen（蒂宾根大学）； Tübingen AI Center（蒂宾根人工智能中心）； HUN-REN Institute for Computer Science and Control (SZTAKI), Budapest, Hungary（匈牙利科学院计算机科学与控制研究所（SZTAKI））

作者：Eric Günther, Balázs Szabados, Kristof Meding, Gunnar König, Sebastian Bordt, Ulrike von Luxburg

英文摘要：Algorithmic explanations are intended to help stakeholders understand opaque algorithmic decisions, but in practice, they often fall short. First, the meaning of algorithmic explanations is often not what one might intuitively expect, so expert knowledge is required to interpret them correctly. Second, recent work has shown that popular explanation algorithms are uninformative about the behavior of complex decision functions. Together, these issues create a gap between what explanations appear to convey and what they actually provide. In this work, we propose Explanation Cards for Explanation Algorithms, which augment standard explanations with complementary information about robustness and validity, as well as clear instructions for interpretation. The complementary information can render otherwise uninformative explanations practically useful, while also helping to detect cases where they are not. Importantly, the interpretation instructions in explanation cards shift responsibility from users to providers: Rather than expecting users to recognize what can and cannot be concluded from an explanation, providers must make this explicit upfront. Using counterfactual explanations and SHAP as examples, we demonstrate how providers can construct explanation cards and that these cards provide users with the guidance needed for sound interpretation. We further argue that explanation cards offer a practical means of operationalising the explainability provisions of the EU AI Act. Overall, explanation cards are a significant step toward making explanation algorithms fit for real-world use cases.

AI 总结：提出一种通过局部区域稳定样本数缩放鲁棒性项的泛化上界，在ImageNet上实现非空洞且最紧的误差估计。

链接：https://arxiv.org/abs/2606.16883

机构：North Carolina Agricultural and Technical State University（北卡罗来纳农业技术州立大学）； University of Alabama（阿拉巴马大学）； Southwest Research Institute（西南研究院）

作者：Abdul-Rauf Nuhu, Parham M. Kebria, Vahid Hemmati, Mahmoud N. Mahmoud, Edward Tunstel, Abdollah Homaifar

英文摘要：Generalization is a critical property of data-driven models, particularly deep learning models deployed in safety-critical applications. Robustness-based generalization bounds have gained attention as a principled way to link robustness properties to generalization performance, often in a data-dependent manner. However, most existing bounds suffer from vacuousness in practical settings, yielding loose upper bounds that greatly exceed the actual error rates and limiting their usefulness for real-world evaluation. While this issue is often attributed to the uncertainty term, a substantial part of the problem originates from the robustness term itself, particularly for the 0-1 loss. Existing approaches typically treat the robustness term as a global measure, ignoring its variation across different sub-regions of the input space. In this work, we propose a generalization bound that addresses this limitation by scaling the robustness term according to the number of stable and unstable samples within each sub-region. Our bounds incorporate both data- and model-dependent factors while maintaining practical relevance (yielding tighter upper bounds on true error). Experiments on models trained on the ImageNet dataset show that our bounds remain consistently non-vacuous and achieve the tightest estimates among existing methods, closely aligning with empirical performance across a range of robust deep neural networks.

AI 总结：本文比较了19种图神经网络层在轨迹预测中的空间和时间处理能力，发现ARMA、Chebyshev和拓扑感知层表现最佳，并总结了基于和聚合、多头注意力和不同跳距权重等设计原则。

链接：https://arxiv.org/abs/2606.14956

机构：Ontario Tech University（安大略理工大学）； Assiut University（艾斯尤特大学）

作者：George Daoud, Mohamed El-Darieby

英文摘要：Autonomous driving systems rely on precise trajectory prediction to plan safe and efficient movement. Graph Neural Networks (GNNs) have become a promising approach for modelling spatiotemporal interactions among road agents. However, designing GNN architectures for trajectory prediction remains non-standardized, with little guidance on which graph layers effectively capture spatial interactions and temporal dynamics. This paper offers a detailed comparative study of 19 graph layer types, focusing on their spatial and temporal processing capabilities to discover the most effective architectures for trajectory prediction. Within the explored hyperparameter setting, we highlight five standout layer combinations, with ARMA, Chebyshev, and topology-aware layers consistently performing better than others. Beyond performance metrics, our findings yield practical design principles: sum-based aggregation is more effective than mean-based methods, multi-head attention mechanisms enable richer interactions, and assigning different weights to different hop distances significantly improves prediction accuracy. These findings offer useful guidance for designing more interpretable and effective trajectory prediction models.

AI 总结：本文形式化了大语言模型处理文本属性图时因图线性化导致的结构失真机制，并提出轻量级推理时修改方法GaLA，通过校正注意力偏差提升零样本图推理性能。

链接：https://arxiv.org/abs/2606.15633

机构：University of Michigan（密歇根大学）； Amazon（亚马逊）

作者：Donald Loveland, Puja Trivedi, Ari Weinstein, Edward W Huang, Danai Koutra

英文摘要：Large Language Models (LLMs) have shown promise for reasoning over Text-Attributed Graphs (TAGs). However, applying LLMs to graphs requires linearizing their structure into sequences, introducing distortion rooted in the graph bandwidth problem. While this distortion has been shown to degrade performance, it is often attributed to prompt design or model scale, leaving the underlying mechanism unclear. In this work, we show \textit{how} rotary positional embeddings turn graph linearization into bandwidth-dependent attention decay, suppressing attention between graph-adjacent nodes that are forced far apart in the serialized sequence. This shifts the focus of LLM-based graph reasoning from prompt engineering and scaling toward correcting attention misalignment. Motivated by this analysis, we propose \textbf{G}raph-\textbf{a}ligned \textbf{L}anguage \textbf{A}ttention (\textbf{GaLA}), a lightweight, inference-time modification for LLMs. GaLA biases attention toward graph-adjacent nodes while preserving the LLM's sequential inductive biases. Across TAG benchmarks, GaLA improves performance with negligible overhead, demonstrating that distortion is a correctable bottleneck in LLM-based graph reasoning.

AI 总结：提出TCHG框架，将信任证据分解为三个通道（实体可靠性、交互行为可靠性、上下文信任），分别控制图传播中的消息准入、传播强度和模式选择，并采用非均匀衰减的时间状态处理多尺度演化，实现可靠动态信任预测。

链接：https://arxiv.org/abs/2606.16611

机构：Xidian University（西安电子科技大学）； Tsinghua University（清华大学）

作者：Bohao Liao, Boyu Deng, Qipeng Song, Jieling Wang, Jingchao Wang

英文摘要：Trust prediction infers latent user-user trust relations and provides important support for social recommendation, fake-review and manipulation detection, and risk identification. Graph neural networks have become a prominent approach to trust prediction because of their ability to learn network structures and complex trust dependencies. However, existing methods often rely on a unified representation of trust signals and do not disentangle heterogeneous trust evidence into separate evidence channels, failing to exploit the distinct roles that different evidence channels should play during trust modeling. To address this gap, this paper argues that trust evidence should not be treated as an undifferentiated input, but should be decomposed and used as functional control factors over graph propagation. We propose TCHG, a tri-trust conditioned heterogeneous graph learning framework that decomposes trust evidence into three channels and assigns them distinct functional roles in propagation: entity reliability governs message admission, interaction-behavior reliability modulates propagation strength, and contextual trust adjusts the propagation mode through context-conditioned operator selection. Since the three evidence channels evolve at different temporal scales, TCHG maintains independent temporal states with non-uniform decay rates to prevent rapidly changing contextual signals from overwriting slowly accumulated entity reliability. It further predicts trust probability and calibrates the output probability, improving predictive confidence under sparse or conflicting evidence. Extensive experiments on multiple public trust datasets show that TCHG achieves effective and reliable trust prediction compared with representative trust prediction and heterogeneous graph baselines.

AI 总结：提出用贝蒂数、谱间隙和解析挠率三个不变量的紧凑谱表示替代全谱，在多个数据集上实现同等或更优性能，显著降低计算开销并避免高频噪声。

链接：https://arxiv.org/abs/2606.16990

机构：University of Southern California（南加州大学）

作者：Jernej Grlj, Aaron D. Lauda

英文摘要：While persistent Laplacians (PL) offer a richer geometric representation of data than persistent homology, utilizing their full eigenspectrum for learning tasks is often hampered by high dimensionality and the ``varying length'' problem across different filtration scales. We propose a compact spectral representation that distills the persistent Laplacian into three mathematically grounded invariants: Betti numbers, the spectral gap, and analytic torsion. Across benchmark datasets including MNIST, QM-3D, and SKEMPI WT, we demonstrate that this reduced feature space captures the essential predictive signal of the full spectrum, and in some cases outperforms it, while significantly reducing computational overhead and preventing the noise introduced by higher-frequency eigenvalues. Our results suggest that these invariants provide a principled, fixed-length interface between spectral geometry and topological learning.

AI 总结：提出GRASP方法，通过序列处理、参数梯度对齐和迭代微调，在O(1)内存下实现多源知识融合，在三个持续学习基准上平均准确率93.5%，优于集成方法的71.7%。

链接：https://arxiv.org/abs/2606.14900

机构：San Diego State University（圣地亚哥州立大学）； University of Utah（犹他大学）； University of Maine（缅因大学）

作者：Mary Isabelle Wisell, Nicholas Jacobs, Aayush Manandhar, Salimeh Yasaei Sekeh

英文摘要：Multi-source transfer learning faces a fundamental scalability bottleneck: existing approaches require either loading all K source models into memory simultaneously during parameter fusion, requiring O(K) memory, or deploying all models at inference time, making production deployment infeasible. We propose GRASP (Gradient-Aligned Sequential Parameter Transfer), which achieves superior knowledge integration while maintaining O(1) memory consumption through three key innovations: (1) sequential processing that merges one source at a time into an evolving target model, (2) parameter-wise gradient alignment that selectively transfers only parameters whose optimization directions align with the target domain, avoiding negative transfer, and (3) iterative fine-tuning that adapts transferred knowledge before integrating the next source. Extensive experiments across three continual learning benchmarks (Yearbook, CLEAR-10, CLEAR-100) spanning 10 to 108-year temporal distribution shifts and four architectures (1.3M to 25.6M parameters) demonstrate that GRASP achieves 93.5% mean accuracy over all datasets and architectures compared to ensemble method's 71.7% accuracy while requiring only constant memory versus K models for standard multi-source fusion. Critically, GRASP's sequential previously merged models and scales to arbitrarily many sources without memory growth, making it uniquely suitable for resource-constrained deployment and continually evolving source domains.

AI 总结：提出跨设备神经算子框架，将平衡重建转化为算子学习问题，通过多几何预训练实现数据高效迁移，Wavelet Neural Operator在100个目标样本下达到低于4%的L2误差。

链接：https://arxiv.org/abs/2606.15512

机构：Grainger College of Engineering, Nuclear, Plasma & Radiological Engineering Department, University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校格兰杰工程学院核、等离子体与放射工程系）； Fusion Energy Division, Oak Ridge National Lab（橡树岭国家实验室聚变能源部）； National Center for Supercomputing Applications（国家超级计算应用中心）； Department of Applied Mechanics, Indian Institute of Technology Delhi（印度理工学院德里分校应用力学系）； Yardi School of Artificial Intelligence, Indian Institute of Technology Delhi（印度理工学院德里分校亚迪人工智能学院）

作者：Jay Phil Yoo, William Howes, Yashika Ghai, Kazuma Kobayashi, Souvik Chakraborty, Syed Bahauddin Alam

英文摘要：Real-time reconstruction of magnetohydrodynamic equilibria is essential for plasma shaping, stability assessment and feedback control in magnetic confinement fusion. However, Grad-Shafranov equilibrium calculations remain largely device-specific and iterative, limiting their use in latency-constrained control settings. Existing neural approaches can accelerate individual equilibrium predictions, but they do not generally provide reusable models across changing plasma boundaries or tokamak geometries. Here we show that equilibrium reconstruction can be recast as a cross-device operator learning problem. We develop a domain-specific neural operator framework that maps geometry and profile parameters directly to the poloidal flux field, replacing repeated solve-on-demand computation with amortized operator inference. Using the analytically tractable Solov'ev family as a controlled Grad-Shafranov testbed, we generate equilibria across eight geometrically distinct tokamak-like configurations and benchmark five neural operator architectures under four transfer-learning strategies. Single-geometry pretraining gives poor transfer to unseen devices, whereas multi-geometry pretraining enables data-efficient adaptation. The Wavelet Neural Operator gives the strongest cross-geometry performance, reaching mean relative L2 errors below 4% with 100 labelled target equilibria and below 2% with full fine-tuning. The predicted magnetic fields satisfy the divergence-free constraint to numerical precision, and four architectures achieve millisecond or sub-millisecond inference. These results identify neural operator pretraining as a route towards reusable, real-time equilibrium inference across fusion device configurations.

AI 总结：研究后训练各阶段（CPT、SFT、RL）对生物学推理模型领域内和领域外性能的影响，发现SFT提升领域内性能但损害泛化，RL可部分恢复泛化，最佳策略是短SFT加长RL。

链接：https://arxiv.org/abs/2606.16517

机构：Harvard University（哈佛大学）； Google DeepMind（谷歌DeepMind）； Google Research（谷歌研究院）

作者：Lukas Fesser, Hanlin Zhang, Michelle M. Li, Eric Wang, Bryan Perozzi, Shekoofeh Azizi, Sham M. Kakade, Marinka Zitnik

英文摘要：Scientific reasoning models for biology combine language models with foundation models trained on multimodal biological data, including DNA, RNA, and proteins. These models are built through post-training, yet how each stage shapes reasoning and generalization remains poorly understood. We study when post-training improves performance and when it induces over-specialization. Across genomics, transcriptomics, and proteins, we train and evaluate more than 100 biological reasoning models under controlled variation in backbone, continued pre-training (CPT), supervised fine-tuning (SFT), and reinforcement learning (RL), measuring both in-domain (ID) and out-of-domain (OOD) performance. We find that each post-training stage reshapes generalization in a distinct way rather than contributing uniform gains. CPT improves downstream performance by aligning models with biological language. SFT consistently increases ID performance but causes OOD performance to peak early and decline as models fit the training distribution. RL, when applied to strong SFT checkpoints with aligned rewards, improves OOD performance and partially recovers generalization. These results show that biological reasoning does not improve monotonically with additional supervision or compute. Instead, performance depends on how training stages are composed. Under fixed post-training budgets, the strongest ID-OOD trade-off comes from brief SFT, larger RL allocations, and asymmetric adaptation capacity across stages.

AI 总结：提出CILN框架，通过受控输入扰动生成实例相关标签噪声，构建90个基准设置，揭示噪声结构对算法行为的重要影响。

链接：https://arxiv.org/abs/2606.14965

作者：Shadman Islam, Agustinus Kristiadi, Mostafa Milani

英文摘要：Synthetic instance-dependent label noise (IDN) benchmarks are widely used to evaluate noisy-label learning methods, yet existing approaches typically generate noise through imperfect annotators or classifier raters, leaving the source of ambiguity implicit. We introduce CILN, a benchmark generation framework that creates IDN through controlled input corruptions. A diverse voter pool labels corrupted instances, producing benchmark datasets in which both the source and severity of ambiguity are explicit and controllable. Using CIFAR10, MNIST, and Adult, we construct 90 benchmark settings spanning multiple corruption families and severity levels. Our experiments show that the resulting benchmarks exhibit genuine instance-dependent noise, provide diverse confusion structures, and, on CIFAR-10, can produce label distributions that are closer to human uncertainty than an existing synthetic IDN benchmark. We further demonstrate that corruption-mediated IDN can expose failure modes of popular noisy-label learning methods, including Co-Teaching and DivideMix, that are not observed under comparable levels of rater-fallibility noise. These findings suggest that noise structure, not only noise rate, plays an important role in benchmark difficulty and algorithm behavior. By making ambiguity generation explicit and controllable, CILN provides a complementary benchmarking framework for studying noisy-label learning under diverse sources of instance difficulty.

AI 总结：提出FastMix框架，将数据混合选择重新表述为双层优化问题，通过联合优化混合系数和模型参数，实现高效、可扩展的数据混合发现，在预训练和后训练中均优于基线方法且大幅降低搜索成本。

链接：https://arxiv.org/abs/2606.14971

机构：University of Hong Kong（香港大学）； Tencent（腾讯）； Chinese University of Hong Kong（香港中文大学）

作者：Haoru Tan, Sitong Wu, Yanfeng Chen, Jun Xia, Ruobing Xie, Bin Xia, Xingwu Sun, Xiaojuan Qi

英文摘要： While large and diverse datasets have driven recent advances in large models, identifying the optimal data mixture for pre-training and post-training remains a significant open problem. We address this challenge with FASTMIX, a novel framework that automates data mixture discovery while training only a single proxy model. Instead of relying on predefined heuristics or resource-intensive simulations, FASTMIX jointly optimizes mixture coefficients and model parameters, substantially improving efficiency and scalability over prior approaches. At the core of FASTMIX is a reformulation of mixture selection as a bilevel optimization problem. Under this reformulation, we show that optimizing mixture ratios is mathematically equivalent to assigning per-source loss weights under uniform source sampling. This embeds the mixture coefficients directly into the differentiable iterative optimization objective, enabling efficient, gradient-based optimization of both mixture and model. To solve the optimization problem, FASTMIX implements an approximate iterative optimization procedure, alternating between (i) updating model parameters on data sampled according to current mixture ratios (inner loop) and (ii) updating mixture ratios based on validation feedback (outer loop). Across pre- and post-training, FASTMIX outperforms baselines while drastically reducing search cost. Code ( this https URL )

AI 总结：本文指出世界模型评估中声明与证据不匹配的问题，提出以决策为中心的评估框架，强调反事实推理、策略优化等能力，并定义L0-L7评估阶梯。

链接：https://arxiv.org/abs/2606.15032

机构：National Key Laboratory for Novel Software Technology, Nanjing University（南京大学计算机软件新技术国家重点实验室）； School of Artificial Intelligence, Nanjing University（南京大学人工智能学院）； Cirquar Technologies

作者：Yang Yu, Shiyuan Zhang, Yifei Sheng, Haoxiang Ren, Haoxin Lin

英文摘要：World models have rapidly become one of the central abstractions in modern AI. Yet the term now refers to several different objects: action-conditioned environment models, latent imagination models, future-video predictors, interactive neural simulators, latent predictive representations, and synthetic-data engines. Evaluation has broadened with the term. Recent papers measure video realism, perceptual similarity, instruction following, physical plausibility, policy ranking, executability, planning success, and downstream policy improvement. The result is not only metric diversity but also a recurring problem of claim/evidence mismatch: papers frequently make a stronger claim about what their model is useful for than their evaluation can actually establish. This paper surveys the recent literature and argues that the central question is use-dependent. When a model is presented as a world model for embodied decision-making, a more decisive issue is not whether it generates visually compelling videos, but whether it supports reliable counterfactual reasoning, policy evaluation, planning, and policy optimization under intervention, policy-induced distribution shift, and long-horizon rollout. We organize the literature using an L0--L7 ladder that ranges from visual plausibility to policy optimization utility. In our interpretation, L0--L3 are most naturally read as diagnostics of generated artifacts, L4 is often the first genuinely interventional test, and L5--L7 provide the most direct evidence of decision usefulness. Based on this diagnosis, we propose a decision-making-centric evaluation framework and a benchmark protocol that foreground counterfactual action fidelity, closed-loop rollout validity, reward/value prediction, policy-ranking agreement, optimization lift, model exploitability, and uncertainty calibration.

AI 总结：针对现有船舶轨迹预测基准缺乏统一协议和环境上下文的问题，提出EnvShip-Bench，基于丹麦海事局和NOAA的原始AIS数据构建，采用标准化预测协议，提供环境与邻近船舶上下文扩展，支持轨迹、环境感知和交互感知预测的统一评估。

链接：https://arxiv.org/abs/2606.15240

机构：Harbin Engineering University（哈尔滨工程大学）； Politecnico di Torino（都灵理工大学）； Institute of Software, Chinese Academy of Sciences（中国科学院软件研究所）

作者：Kun Ma, Qilong Han, Chengjing Song, Jingzheng Yao, Hao Wang, Changmao Wu

英文摘要：Vessel trajectory prediction is important for intelligent shipping, maritime surveillance, and navigation safety. However, existing public maritime AIS resources are often limited by inconsistent forecasting protocols, uneven data quality, and the lack of benchmark-ready contextual annotations, which hinder fair comparison and context-aware modeling. To address this gap, we present EnvShip-Bench, a unified benchmark for short-term vessel trajectory prediction built from large-scale raw AIS data from the Danish Maritime Authority (DMA) and NOAA through a common processing pipeline. EnvShip-Bench adopts a standardized forecasting protocol with 10 minutes of observation, 10 minutes of prediction, and 20-second sampling in vessel-centric local metric coordinates. Beyond the large-scale core benchmark, it provides a quality-first compact subset for efficient and reproducible experimentation, together with synchronized environmental and nearby-vessel context extensions. As a result, EnvShip-Bench supports trajectory-only, environment-aware, and interaction-aware forecasting under a unified evaluation framework. Extensive benchmark statistics and analysis demonstrate that EnvShip-Bench offers a standardized, extensible, and context-aware foundation for maritime trajectory forecasting research.

AI 总结：提出LatentGym测试平台，通过可控潜在变量分离探索与利用，研究LLM代理在跨任务序列中的适应性学习机制。

链接：https://arxiv.org/abs/2606.15306

作者：Daksh Mittal, Tommaso Castellani, Thomson Yen, Naimeng Ye, Fangyu Wu, Minghui Chen, Tiffany Cai, Emmanouil Koukoumidis, William Zeng, Hongseok Namkoong

英文摘要： We envision continually learning agentic systems that become more useful over time: as they encounter sequences of related tasks, they should infer the hidden structure shared across those tasks and use it to improve future decisions. This cross-task experiential learning capability is pivotal in domains such as personalization and interactive assistance, but existing training/evaluation frameworks do not provide shared, controllable latent structures and cannot measure whether or why agents improve. We introduce LatentGym: a controllable suite in which each environment is organized around a ground-truth latent variable governing the structure across tasks. Our construction yields metrics that separate exploration (whether the agent's actions gather information about the latent) from exploitation (whether the agent uses what it has gathered). We demonstrate our suite on empirical studies addressing three questions: how and why frontier models fail to adapt across related tasks; whether post-training on related task sequences improves general cross-task adaptation, and where those gains come from; and how design choices such as inter-task feedback shape training dynamics and generalization. Together, these results establish a controlled foundation for studying how LLM agents learn from experience across tasks, and for designing agents that adapt more reliably in sequential, personalized, and interactive settings.

AI 总结：提出多模型多目标咳嗽回归基准，评估五个基础模型在六个目标上的表现，发现MLP-small优于线性探测，揭示数据集大小与头部容量的权衡，并展示跨数据集迁移的不对称性。

链接：https://arxiv.org/abs/2606.15436

作者：Mayur Sanap, Prasanna Desikan, Edgar Lobaton

英文摘要：Respiratory acoustic foundation models (FMs) excel at cough classification, yet their ability to predict continuous health quantities from cough audio remains largely unexplored, despite the clinical value of passive age, BMI, and disease probability estimation in settings where physical measurements are unavailable. We introduce the multi-model, multi-target cough regression benchmark evaluating five FMs (OPERA-CT, OPERA-CE, OPERA-GT, HeAR, M2D+Resp) across six targets on three datasets under subject-disjoint protocols, comparing linear, MLP-small, and full MLP regression heads. MLP-small beats the mean-predictor baseline on all tasks and linear probing in 23 of 30 model x task cases, with full MLP overfitting on small clinical data but recovering on larger sets, revealing a dataset size x head-capacity trade-off. HeAR leads within-dataset age regression on Coswara (9.12 yr MAE); its CIDRZ result is excluded from headline claims owing to possible HeAR-CIDRZ pretraining overlap. OPERA-GT is favored over OPERA-CT on age in all three datasets, with the CIDRZ margin within seed variance, extending a generative-pretraining advantage from breath to cough. HeAR and M2D+Resp reach near-full performance at N = 50 samples while OPERA models require N = 400. Cross-dataset transfer is strongly asymmetric as large diverse data generalises to small clinical populations (CoughVID to CIDRZ: -0.17 yr) but not vice versa (CIDRZ to Coswara: +2.43 yr, +26.6%).

AI 总结：通过分离中间表示与执行机制，在40个任务上比较代码执行与自然语言推理，发现代码执行优势源于外部执行而非表示变化。

链接：https://arxiv.org/abs/2606.15589

作者：Terry Tong, Yu Feng, Surbhi Goel, Dan Roth

英文摘要：For tool-augmented language models, comparing natural-language reasoning with code-execution pipelines is difficult because the comparison changes both the intermediate representation and the execution mechanism. We separate these factors with an intermediate intervention: the model expresses its reasoning as executable code, and the language model simulates that code in context to produce an answer. On a 40-task verifiable algorithmic benchmark, deterministic code execution outperforms natural-language reasoning by +31.6pp. We observe that the intermediate intervention is not meaningfully different from natural-language reasoning (+0.15pp). These results suggest that, in our evaluated setting, changing the intermediate representation alone does not explain the tool-use advantage, providing evidence for the performance gains requiring reliable external execution. We formalize this intuition with a simple statistical decision-theoretic model that characterizes when execution dominates end-to-end risk in our disentangled trace-generation/execution regime. We validate our theory using a reconstruction intervention that leverages a proxy language model to infer natural-language reasoning traces from code representations, recovering performance comparable to the original natural-language reasoning pipeline. All experiments are at this https URL.

AI 总结：通过三遍实验设计，测量了在反事实令牌信用估计中重新喂食前缀导致的噪声，发现其改变信用估计的比率高于副本噪声基底，建议恢复解码器状态或使用批不变内核。

链接：https://arxiv.org/abs/2606.15621

机构：Northeastern University（东北大学）

作者：Nils Matteson

英文摘要：Per-token counterfactual credit estimation asks which token in a language-model rollout caused the final answer to be right or wrong: cut the transcript at a pivot, substitute an alternative token, replay continuations, and compare outcomes. Published methods re-feed the transcript prefix as a fresh prompt, assuming this reproduces the state the model passed through during generation. We measure what that assumption costs on a stock inference engine, with a three-pass design: continuations resumed from the verified decode-time KV state, an identical second exact pass (a replica noise floor), and a re-feed pass. Across six configurations and three models (including a GRPO-trained checkpoint), at low-margin decision tokens, re-feeding changes the credit estimate at rates 14-28 percentage points above the replica floor (7-21pp under a treatment-independent conditioning; problem-clustered t = 2.9-6.4). Most changes are zero-boundary crossings of the quantized estimator rather than polarity reversals, and the perturbation is consistent with mean-zero, so averaged quantities are largely safe; but selection is not: a critical-token set chosen by thresholding $|\hat{A}_t|$ under re-feed overlaps the exact-resume selection at Jaccard 0.34-0.90, versus a 0.63-0.96 replica ceiling. A causal confirmation closes the loop: under vLLM's batch-invariant kernels all three passes are identical on every measured channel, with both disagreement rates exactly zero. Replica passes themselves disagree on 9-23% of eligible estimates: single-sample credit measurements at decision tokens are unreliable under any replay. Settings were fixed in advance; exact-pass cache hits in the second campaign are instrumented (100% hit rate, 3,434 pivots); total compute was under 10 USD. We recommend that counterfactual credit studies resume decoder state or use batch-invariant kernels, and report a replica floor.

AI 总结：针对深度学习理论与实践的差距，提出一个基准框架，通过扩展dSprites和COIL-20数据集并配合有限差分估计器，实现曲率、可达性和体积的近真实值估计，用于校准几何估计器和验证理论假设。

链接：https://arxiv.org/abs/2606.15760

作者：Marios Koulakis, Constantin Seibold

英文摘要：A significant gap exists between theory and practice in deep learning. Generalization and approximation error bounds are often derived for simplified models or are too loose to be informative. Many rely on the manifold hypothesis and on geometric regularity such as intrinsic dimension, curvature, and reach. Progress requires insight into data-manifold geometry and suitable benchmarks, yet existing options are polarized: analytic manifolds with known geometry but limited applicability, or real-world datasets where geometry is only coarsely estimable. We introduce a benchmarking framework for studying data geometry. We repurpose and extend dSprites and COIL-20 with additional transformation dimensions and dense, axis-aligned sampling, and pair them with finite-difference estimators that recover curvature, reach, and volume at near-ground-truth accuracy in a regime where general-purpose estimators are unreliable or difficult to deploy. The framework is intended as a controlled testbed, useful as a calibration environment for geometric estimators and a sandbox for probing theoretical assumptions. To illustrate its use, we present two application studies, namely assessing the scaling behavior of the bounds of Genovese et al. and Fefferman et al., and tracking the layer-wise geometry of a $\beta$-VAE, highlighting the behavior of current bounds and the value of controlled benchmarks for guiding and validating future theory. A reference implementation is available at this https URL.

AI 总结：本文通过系统基准测试，比较了简单计数 argmax 基线、LSTM、Transformer 和 LLM 在下一活动预测中的性能，发现 argmax 基线在多数数据集上可媲美或接近十亿参数 LLM。

链接：https://arxiv.org/abs/2606.15868

机构：Technical University of Munich（慕尼黑工业大学）； Fraunhofer Gesellschaft（弗劳恩霍夫协会）

作者：Hans Weytjens, Ingo Weber

英文摘要：Next activity prediction (NAP) is a cornerstone of predictive process monitoring (PPM), enabling organizations to move from retrospective analysis to proactive process steering. The PPM field has progressed from classical machine learning through deep learning architectures such as LSTMs and Transformers to large language models (LLMs). Despite growing model complexity, no benchmark jointly compares LLMs, Transformers, LSTMs, and simple baselines in a direct sequence modeling setting for NAP. In this paper, we fill this gap with a systematic benchmark. We compare vocabulary-adapted LLMs, Transformers trained from scratch, LLM-distilled Transformers, and LSTMs against a simple counting-based argmax baseline across seven real-life event logs. Our results tell a David vs. Goliath story: pretraining confers no consistent improvement over training from scratch, model size shows little effect on performance, and on most datasets the argmax baseline matches or approaches the performance of billion-parameter LLMs.

AI 总结：本研究验证了LLM系统AIPR通过提示对论文进行评分，无需微调，其整体评分能有效区分ICLR会议的接收与拒绝论文（AUROC 0.82），且评分稳定、可复现，为辅助同行评审提供了可靠依据。

链接：https://arxiv.org/abs/2606.15887

作者：Costa Georgantas

英文摘要：Large language model (LLM) systems are increasingly proposed to assist peer review, yet most evaluations judge the prose of machine-generated review text, not the validity of the numeric score a system assigns. We validate AIPR, which reads a submitted manuscript and emits five 0-100 quality dimensions and a weighted overall score, against the public decision outcomes of a major machine learning venue. AIPR grades by prompting alone, with no fine-tuning on reviews or decisions. Across 300 ICLR submissions with public decision tiers and reviewer ratings, graded under a frozen pipeline with hypotheses pre-registered before any score met any outcome, the overall score separates rejected from accepted submissions (AUROC 0.82, 95% CI 0.78-0.87), rises monotonically across tiers, and tracks the mean reviewer rating. The signal is strongest where we claim it: the lowest-scoring fifth is rejected far above the base rate, with oral papers absent. The validity comes mostly from the model: a one-paragraph prompt on the same model discriminates almost as well as the full pipeline (the small gap favours the pipeline but does not meet the pre-declared criterion, p = 0.09). What the engineering adds is reliability and a grounded review: AIPR's score barely moves across repeated runs (0.7 vs. 2.8 points within-paper SD) where the bare prompt swings, and the same pass returns a rubric-structured, evidence-grounded review rather than a bare number, with the human keeping the decision.

AI 总结：提出基于低秩近似和残差采样的数据选择框架，在温和正则条件下选择加权子集，使平均损失近似全数据集平均损失，相对误差(1+ε)加性项εΦ_k，实验优于均匀采样和聚类敏感采样。

链接：https://arxiv.org/abs/2606.16045

机构：Google Research（谷歌研究院）； University of California, Berkeley（加州大学伯克利分校）； Institute of Science and Technology Austria (ISTA)（奥地利科学技术研究所）； Carnegie Mellon University（卡内基梅隆大学）； Texas A&M University（德克萨斯农工大学）

作者： Vincent Cohen-Addad, Sasidhar Kunapuli, Vahab Mirrokni, Mahdi Nikdan, David P. Woodruff, Samson Zhou

英文摘要：In the data selection problem, the objective is to choose a small, representative subset of data that can be used to efficiently train a machine learning model. Sener and Savarese [ICLR 2018] showed that, given an embedding representation of the data and suitable geometric assumptions, heuristics based on $k$-center clustering can be used to perform data selection. This perspective was further explored by Axiotis et. al. [ICML 2024], who proposed a data selection approach based on $k$-means clustering and sensitivity sampling. However, these methods rely on the assumption that the dataset exhibits intrinsic geometric structure that can be effectively captured by clustering, whereas many modern datasets instead possess global algebraic structure that is better exploited by low-rank approximation or principal component analysis. In this paper, we introduce a new data selection framework based on low-rank approximation and residual-based sampling, formulated through the lens of row subset selection and loss-preserving coreset construction. Given an embedding representation of the data satisfying mild regularity conditions, which can be interpreted as algebraic or angular notions of Lipschitz continuity, we show that it is possible to select a weighted subset of $\tilde{O}\left(k + \frac{1}{\varepsilon^2}\right)$ data points whose average loss approximates the average loss over the full dataset within a $(1+\varepsilon)$ relative error, up to an additive $\varepsilon \Phi_k$ term, where $\Phi_k$ denotes the optimal rank-$k$ approximation cost of the embedding matrix. We complement these theoretical guarantees with empirical evaluations, demonstrating that on a range of real-world datasets, our data selection approach achieves improved performance over prior strategies based on uniform sampling or clustering-based sensitivity sampling.

AI 总结：针对数据受限下标准自回归预训练严重过拟合的问题，提出三类数据增强方法（token级噪声、序列排列、目标偏移预测），有效降低验证损失并支持数百epoch训练。

链接：https://arxiv.org/abs/2606.16246

机构：UC San Diego（加州大学圣地亚哥分校）； RMIT University（皇家墨尔本理工大学）

作者：Michael K. Chen, Xikun Zhang, Zhen Wang

英文摘要：As AI labs approach a data ceiling where compute capacity outpaces the rate of new high-quality text generation, language model pretraining is shifting toward a data-constrained, compute-abundant regime that demands productive multi-epoch training on fixed corpora. Standard autoregressive (AR) pretraining overfits severely in this setting, reaching its optimum early and then continuously deteriorating. We investigate data augmentation as a regularizer to mitigate this overfitting and enable productive training for hundreds of epochs on the same data. We introduce three orthogonal categories of augmentation for AR pretraining: token-level noise (masking, random replacement), sequence permutations (right-to-left prediction, Fill-in-the-Middle), and target offset prediction ($x_{t+i}$ for $i > 1$). Through systematic ablations, we find that individual augmentations delay overfitting and lower validation loss relative to the baseline, with random token replacement achieving the best minimum loss among individual methods. Combining augmentation categories further lowers the minimum validation loss. Our experiments demonstrate that data augmentations mitigate AR pretraining's data inefficiency and offer a promising solution to the data-constrained regime. All code and data are available at this https URL

AI 总结：本文研究过滤式近似最近邻搜索中，选择性估计误差如何导致计划遗憾，并揭示其仅在相变边界附近产生，遗憾呈对数宽度楔形，通过有限尺度标度验证。

链接：https://arxiv.org/abs/2606.16341

机构：VaidhyaMegha Private Limited, India（VaidhyaMegha 私人有限公司，印度）

作者：Madhulatha Mandarapu, Sandeep Kunkunuru

英文摘要：A filtered approximate-nearest-neighbor (ANN) query returns the k nearest vectors among those satisfying an attribute predicate P of selectivity s. The best execution strategy -- pre-filter, post-filter, or in-filter -- changes with s, so a system must estimate s and choose. We model this as an argmax over a landscape with phases (regions where each strategy wins) separated by boundaries, and show that selectivity-estimation error produces plan regret -- recall lost versus the oracle strategy -- only in the critical regions around those boundaries. The regret is a wedge of log-width equal to the multiplicative estimation error epsilon and height equal to the local cliff |V'(s*)| epsilon; the flip-margin 1/|V'(s*)| is the condition number of a sibling cardinality-estimation study reappearing as the local boundary theory. The two phase boundaries follow from independent mathematics: order statistics place the post-filter cliff at s ~ k/K, and site percolation places the in-filter cliff at s_c ~ 0.83/M for graph degree M (corpus-size independent). Criticality exists only under a constrained budget B < sqrt(k n). Under pre-registered decision rules we confirm, on synthetic sweeps and real SIFT1M, that regret concentrates ~290x at the boundary and that the regret curves obey a finite-size scaling collapse onto one universal wedge across two decades of corpus size. A real approximate index does not mis-locate the boundary, but a biased cost model opens a persistent miscalibration band that estimation-error robustness cannot fix. The contribution is a characterization, not a new index. Code and the full pre-registration are public.

AI 总结：提出SA-MSCP方法，通过块自助法从交叉验证残差生成未来路径并构建经验分位数预测区间，提升聚合和增长率目标的经验覆盖率。

链接：https://arxiv.org/abs/2606.16356

作者：Andro Sabashvili

英文摘要： We study uncertainty quantification for aggregated forecasting tasks such as annual totals and year-over-year growth rates. We propose SA-MSCP, a simulation-augmented multi-step split conformal method that generates future paths from cross-validated residuals using a block bootstrap and constructs prediction intervals from empirical quantiles. Experiments show that SA-MSCP improves empirical coverage over a simulated-path baseline for aggregated and growth-rate targets. Our results demonstrate that simulation-enhanced conformal calibration is an effective and general framework for uncertainty quantification in aggregated time-series forecasting.

AI 总结：针对合成表格数据保真度评估中 Jensen-Shannon 散度估计协议不明确的问题，系统研究了不同估计器族、采样协议等因素对估计值的影响，揭示了边际估计器的依赖盲性和分类器估计器的敏感性，并提出了后验校正方法。

链接：https://arxiv.org/abs/2606.16411

机构：Information Processing and Telecommunications Center, ETSI Telecomunicación, Universidad Politécnica de Madrid（马德里理工大学电信工程学院信息处理与电信中心）

作者：Alba Garrido, Alejandro Almodóvar, Mar Elizo, Patricia A. Apellániz, Santiago Zazo, Juan Parras

英文摘要：The Jensen-Shannon divergence is widely reported as a scalar measure of fidelity for synthetic tabular data. Yet, in practice, it is estimated from finite samples using protocols that are often underspecified. This creates a measurement problem. Although the population divergence is well defined, the empirical value depends on the estimator family, sampling protocol, calibration, dimensionality, and class balance. We show that different protocols can yield non-comparable values: marginal-based estimators ignore dependencies in the joint distribution and can severely underestimate divergence, while classifier-based estimators capture joint structure but exhibit strong estimator dependence. We systematically study this behavior across controlled settings with reference divergences and real-world synthetic tabular benchmarks. Our analysis reveals dependence blindness in marginal estimators, prior-shift bias under class imbalance, and estimator sensitivity in high dimensions. To address prior shift, we derive a closed-form posterior correction for classifier-based Jensen-Shannon estimation. Our results show that empirical Jensen-Shannon divergence values are inherently protocol-dependent, making explicit specification of the estimation procedure necessary for meaningful comparison. We provide practical guidelines and an open-source tool for estimator-aware Jensen-Shannon evaluation.

AI 总结：本文提出一个协议，用于检验LLM评估中尾部形状估计的假阳性，通过极值理论指标区分尾部重量和尾部质量，并在毒性评估中识别出三种假阳性模式。

链接：https://arxiv.org/abs/2606.16511

机构：Sapienza University of Rome（罗马大学）

作者：Luca Zhou

英文摘要：Recent work motivates moving large language model (LLM) evaluation from mean-based to tail-aware metrics, including conditional value-at-risk and tail-index estimates of reward-model error. We ask whether the canonical extreme-value-theory tail-index parameter, which isolates how heavy a tail is from how large the tail mass is, adds discriminative information beyond the mean and a standard tail-magnitude statistic in LLM evaluation. We pre-register a protocol covering admissibility, goodness-of-fit, threshold-stability, and effect-size requirements for any positive tail-shape claim. The protocol is the contribution of this paper; the empirical study below is a demonstration of what its gates catch. Applied to a standard LLM toxicity-evaluation setup under two structurally different scorer families, the protocol catches three distinct modes of false positives that a naive analysis would have published, and rejects the headline tail-shape claim on both scorers. We conclude that tail-shape estimation in the LLM toxicity-evaluation setups we examined is more fragile than the recent literature suggests, and recommend the protocol as a starting point for tail-index claims in similar setups.

AI 总结：提出MIRAGE基准，包含1200个提示，覆盖直接完成、思维链推理和模拟智能体决策三种部署场景，发现思维链放大偏见、智能体决策存在不对称性、偏见与检索新闻时间耦合，现有缓解措施效果有限。

链接：https://arxiv.org/abs/2606.16562

作者：Noor Islam S. Mohammad, Tamim Sheikh

英文摘要：Five years after the discovery of persistent anti-Muslim bias in large language models, most evaluations remain confined to single-turn prompt completion, a setting that no longer reflects how frontier LLMs are deployed. We introduce \textbf{MIRAGE} (Muslim-Identity Reasoning and Agentic Generation Evaluation), a benchmark of 1{,}200 prompts spanning three deployment-realistic conditions: direct completion, chain-of-thought reasoning, and simulated agentic decision-making across content moderation, lending triage, refugee claim summarization, and hiring screens. Across six frontier models, we find that (i) chain-of-thought reasoning \emph{amplifies} rather than suppresses Muslim-violence associations by 12--34\% relative to direct completion, (ii) agentic decisions exhibit a 9--22 percentage-point asymmetry between Muslim and matched non-Muslim cases on identical evidence, and (iii) bias is sharply time-coupled to retrieved news context, increasing 18--27\% under recent-conflict retrieval. Existing prompt-based mitigations transfer poorly across our three conditions, suppressing direct-completion bias while leaving agentic asymmetry largely intact. We release MIRAGE and an open evaluation harness to support targeted mitigation research.

AI 总结：提出MyPCBench基准，在模拟真实桌面环境（含17个Web应用）中测试个人计算机使用代理，发现最佳模型Claude Opus 4.6仅解决55.4%任务，失败集中在多应用和长轨迹任务。

链接：https://arxiv.org/abs/2606.16748

机构： Carnegie Mellon University（卡内基梅隆大学）

作者：Lawrence Keunho Jang, Andrew Keunwoo Jang, Jing Yu Koh, Ruslan Salakhutdinov

英文摘要：Current benchmarks for computer-use agents evaluate models in impersonal environments. This leaves a gap between evaluation and deployment where personal assistants are expected to work across a user's whole digital life, including their context, historical data, and logged-in accounts. This gap is widest on web tasks, where live web evaluations cannot exercise sites that require logging in or personal information, the kind of site a real personal assistant has to drive. We introduce MyPCBench, which tests computer-use agents as personal assistants on a Linux desktop populated with 17 simulated real-world web applications and a full desktop stack, all seeded for one canonical persona, Michael Scott from The Office. We define 184 tasks in this environment, each inspired by a real request drawn from the OpenClaw community, and benchmark six closed and open-weight models with a uniform computer+bash tool surface. We find that the best model, Claude Opus 4.6, fully solves 55.4\% of the tasks, the only model above 50\%. Model failures cluster on tasks that span many applications and on long trajectories, where personalization stresses an assistant the most. We release the environment, task set, and agent harness at this https URL.

AI 总结：提出一个可复现的流水线，生成雷诺数1000-10000的三维通道流训练数据，使用累积碰撞算子的格子玻尔兹曼求解器，并通过实验测量和网格收敛研究验证，为神经算子标准化比较提供基础。

链接：https://arxiv.org/abs/2606.16765

作者：Lukas Schröder, Shubham Kavane, Harald Köstler

英文摘要：Evaluating neural operators for 3D turbulent flow requires validated datasets with physical benchmarks. We present a reproducible pipeline generating training data for 3D channel flows around generated geometries at Re=1,000-10,000. Our lattice Boltzmann solver with cumulant collision operators is rigorously verified against experimental measurements (Strouhal number, drag coefficients, turbulent fluctuations) with comprehensive grid convergence studies at resolution 1024x512x512. Building upon an established framework, this validated pipeline enables standardized surrogate model comparison. We outline planned systematic evaluation of Fourier Neural Operator and U-Net variants on forecasting, super-resolution, and error correction tasks, using physics-informed metrics to assess turbulent energy cascade representation. Future work will compare computational efficiency between numerical solvers and neural surrogates, exploring practical application. We seek community feedback on our validation approach, planned benchmark methodology, and evaluation priorities for neural operators in turbulent flows.

AI 总结：提出HawkesNest基准，基于多元Hawkes过程定义四个复杂度轴，用于可控测试时空点过程模型在已知结构难度下的性能。

链接：https://arxiv.org/abs/2606.16863

机构：German Research Center for Artificial Intelligence (DFKI), Data Science and its Applications Research Group, Kaiserslautern, Germany（德国人工智能研究中心（DFKI），数据科学及其应用研究组，凯撒斯劳滕）； Department of Computer Science, Rhineland-Palatinate Technical University of Kaiserslautern-Landau (RPTU), Kaiserslautern, Germany（莱茵兰-普法尔茨凯撒斯劳滕-兰道工业大学（RPTU）计算机科学系，凯撒斯劳滕）

作者：Yahya Aalaila, Sumantrak Mukherjee, Gerrit Großmann, Sebastian Vollmer

英文摘要：Evaluation of spatiotemporal point process (STPP) models relies heavily on opaque real-world datasets, where latent generative structure is unknown and model failures are difficult to attribute. We introduce HawkesNest, a generator-aligned benchmark for controlled spatiotemporal pattern complexity built on a multivariate Hawkes backbone. HawkesNest defines four complexity axes: space--time entanglement, background heterogeneity, cross-type interaction, and domain topology. Each axis is associated with a deterministic index computed from the latent data-generating mechanism. By varying these axes while holding global rate, stability, and simulation budget fixed, HawkesNest enables diagnostic stress tests of STPP models under known structural difficulty. We verify that the indices are monotone and nearly orthogonal under controlled sweeps. We illustrate its use by showing that Hawkes-family baselines degrade under joint heterogeneity--entanglement complexity, even though they are structurally aligned with the Hawkes data-generating backbone. We further show that HawkesNest exposes neural-model sensitivity: AutoSTPP remains vulnerable under isolated increases in space--time entanglement. Code. Available at this https URL

AI 总结：提出过滤共形椭球方法，结合状态空间滤波与共形校准，为多元时间序列生成联合预测集，控制单事件并适应跨坐标依赖，通过可观测预测律商分析保证覆盖界。

链接：https://arxiv.org/abs/2606.17014

机构：DRW London（DRW伦敦）

作者：Yannick Limmer

英文摘要： Joint prediction sets for multivariate time series should control a single event while adapting to cross-coordinate dependence. We study filtered conformal ellipsoids: a frozen state-space filter emits a one-step predictive mean and covariance, and split-conformal calibration is applied to the resulting Mahalanobis scores. The filter is used to choose the ellipsoid shape; conformal calibration chooses the scalar radius, so the construction benefits from a learned predictive covariance without relying on Gaussian tail probabilities for coverage. The main difficulty is that filtered scores are dependent and learned recurrent filters need not contract in their raw hidden state; we therefore analyse contraction in an observable predictive-law quotient that identifies hidden states producing the same future sequence of emitted Gaussian laws. Under a stable Bayes Gaussian-projection filter, covariance bounds, and a finite-horizon observability Fisher condition, small excess Gaussian negative log-likelihood implies contraction of the learned emitted laws. Combined with a threshold-autocovariance envelope this yields a Chebyshev-type approximate coverage bound for filtered split-conformal prediction under dependence; a sharper Bernstein-type bound requires an additional geometric-mixing concentration assumption. Under Gaussian oracle realisability we also obtain a near-oracle log-volume comparison within the class of conditionally valid Gaussian ellipsoid rules. We instantiate the framework with a GCN-GRU filter with diagonal-plus-low-rank covariance. On moderate-size graph-native traffic benchmarks (METRLA-$20$ and PEMSBAY-$50$), the learned filter gives sharper at-target ellipsoids than static-covariance and non-filter baselines; at full-graph scale and on non-graph-native datasets, factor and copula baselines can be stronger.

AI 总结：提出α-FISP框架，通过约束优化平衡精算公平与团结公平，参数α实现从纯精算到纯团结的连续定价谱，理论保证且计算可行。

链接：https://arxiv.org/abs/2606.14898

机构：Department of Risk and Insurance, Wisconsin School of Business, University of Wisconsin–Madison（威斯康星大学麦迪逊分校威斯康星商学院风险与保险系）； Department of Information Systems and Operations Management, Warrington College of Business, University of Florida（佛罗里达大学沃灵顿商学院信息系统与运营管理系）

作者：Tianhe Zhang, Xiguang Liu, Peng Shi

英文摘要：Fairness in insurance pricing remains a long-standing and deeply debated puzzle. On one hand, insurers, driven by profitability considerations, set premiums that differentiate across individual risks to achieve actuarial fairness. On the other hand, insurance serves a critical societal function by pooling risks across a population, motivating cross-subsidization among groups to promote solidarity fairness. The tension between these two competing notions of fairness makes insurance pricing inherently complex, particularly in modern settings where granular data allow for increasingly fine risk differentiation and regulators face growing pressure to protect vulnerable groups. To address this challenge, we propose an $\alpha$-\textbf{F}air \textbf{I}ndividual \textbf{S}olvent \textbf{P}remium ($\alpha$-FISP) framework for insurance pricing that explicitly captures the trade-off between actuarial and solidarity fairness while guaranteeing solvency, a fundamental requirement in insurance operations. We formulate the pricing problem as a constrained optimization task, where actuarially fair premiums are adjusted subject to budget constraints on cross-subsidization within each risk class. This formulation naturally yields a family of solutions parameterized by $\alpha$, tracing a continuum between purely actuarial and purely solidarity-based pricing and enabling decision-makers to select an operating point along this fairness spectrum. We derive theoretical guarantees for the proposed framework. Numerical experiments show that $\alpha$-FISP is computationally tractable and aligns well with the U.S. regulatory regimes featuring heterogeneous state-level fairness requirements.

AI 总结：针对卡车车队，提出一种基于状态监测的预测性维护方法，将磨损状态建模为单调非递减时间序列，通过选取最近观测并转换为表格数据，利用AutoML简化建模，在Scania组件X数据集上降低了成本。

链接：https://arxiv.org/abs/2606.12486

机构：SnT, University of Luxembourg（卢森堡大学SnT）； Scania CV AB（斯堪尼亚商用车公司）

作者：Valeriu Dimidov, Sasan Jafarnejad, Raphaël Frank

英文摘要：Condition-based Predictive Maintenance (PdM) for truck fleets has gained momentum in recent years. This maintenance strategy aims to minimize unplanned downtimes and reduce costs by monitoring the health status of vehicles and taking proactive action based on their condition. However, the implementation of condition-based PdM systems is challenging due to the large volume of data generated by the trucks, the inherent complexity of detecting failures through sensor data and the difficulties in finding cost-effective trade-offs in the solution's implementation. In this paper, we define and validate a condition-based PdM methodology built on the assumption that the wear-and-tear state of the monitored component can be represented as a monotonically non-decreasing time series. It involves selecting only the most recent observations from the time series and transforming them into a tabular format for classification using machine learning (ML) models designed for tabular data. Our results indicate that the proposed methodology reduces costs on the Scania Component X dataset compared to current state-of-the-art (SOTA) approaches, while also simplifying the modeling process through AutoML.

AI 总结：研究使用机器学习模型分析考试期间的生理数据（皮肤电活动、心率、皮肤温度）预测成绩，比较了逻辑回归、随机森林、SVM及LSTM、GRU、Transformer等模型，发现随机森林在效率和可解释性上表现优异，Transformer与LSTM/GRU性能相当。

链接：https://arxiv.org/abs/2606.14960

机构：Middle Tennessee State University（中田纳西州立大学）

作者：Lala Yamazaki, Ramchandra Rimal

英文摘要： This study investigates the application of machine learning models to predict exam outcomes using physiological data collected during examination sessions. Physiological stress indicators, including electrodermal activity, heart rate, and skin temperature, were analyzed to uncover their association with academic performance. A variety of machine learning approaches were employed, ranging from standard models like logistic regression, random forest, and support vector machines to more advanced architectures, including transformers, long short-term memory (LSTM), and gated recurrent unit (GRU) models. This diversity aimed to capture the complex interactions within the data effectively. A key focus was assessing the adaptability of transformers in processing numerical data and evaluating their performance in this novel context. Standard performance metrics, such as accuracy, precision, recall, and F1-score, were used to compare model efficacy. The experimental results demonstrate that while deep learning models generally excel at capturing complex relationships in physiological data, simpler models like random forests can sometimes achieve superior performance while offering computational efficiency and interpretability. Furthermore, transformers demonstrated notable versatility, showcasing performances comparable to those of the LSTM and GRU models. This research underscores the importance of experimenting with a broad class of models that align with the objectives of the problem at hand, balancing precision, efficiency, and interpretability. By elucidating the relationships between physiological signals and academic performance, this study contributes to understanding stressors affecting students' mental health. It further promotes leveraging physiological data to enhance student well-being and academic outcomes.

AI 总结：针对X射线散射数据离线探索和实时分析两大挑战，训练领域特定注意力卷积变分自编码器（C-VAE），学习低维表示以捕捉结构变化，并集成到MLExchange平台的Latent Space Explorer中，支持交互式结构探索。

链接：https://arxiv.org/abs/2606.14999

机构：Advanced Light Source, Lawrence Berkeley National Laboratory（劳伦斯伯克利国家实验室先进光源）； Center for Advanced Mathematics for Energy Research Applications, Lawrence Berkeley National Laboratory（劳伦斯伯克利国家实验室能源研究应用高级数学中心）； Molecular Biophysics & Integrated Bioimaging Division, Lawrence Berkeley National Laboratory（劳伦斯伯克利国家实验室分子生物物理学与综合生物成像部）； Berkeley Synchrotron Infrared Structural Biology program, Lawrence Berkeley National Laboratory（劳伦斯伯克利国家实验室伯克利同步辐射红外结构生物学项目）； Materials Sciences Division, Lawrence Berkeley National Laboratory（劳伦斯伯克利国家实验室材料科学部）； McKetta Department of Chemical Engineering, University of Texas（德克萨斯大学麦凯塔化学工程系）

作者：Monika Choudhary, Xiaoya Chong, Runbo Jiang, Wiebke Koepp, Petrus H. Zwart, Damon English, Gregory M. Su, Eric Schaible, Chenhui Zhu, Mostafa Nassr, Noah P. Wamble, Kelvin Kam-Yun Li, Jonathan M. Chan, Jose Carlos Diaz, Cameron McKay, Lynn Katz, Benny Freeman, Guillaume Freychet, Yevgen Matviychuk, Eliot Gann, Daniel B. Allan, Benedikt Sochor, Frank Schluenzen, Stephan V. Roth, Ethan Crumlin, Dylan McReynolds, Tanny Chavez, Alexander Hexemer

英文摘要：Scientific user facilities generate X-ray scattering data faster than traditional workflows can process them. We address this challenge across two settings, offline dataset exploration and live on-the-fly analysis. We train a domain-specific attention-based Convolutional Variational Autoencoder (C-VAE) on 1.5 million X-ray scattering images to learn low-dimensional representations capturing structural variation across diverse experimental conditions. The learned latent space reveals well-organized clusters and smooth trajectories reflecting experimental progression. It further supports controlled synthetic scattering image generation across diverse structural states. When deployed without retraining, the model organizes time-resolved film formation experiments at two synchrotron facilities into interpretable latent structures. Benchmarking against DINOv3 (ViT-7B), a general-purpose vision foundation model, demonstrates that domain-specific training yields more interpretable latent organization for scattering data. Both workflows are integrated within Latent Space Explorer, a component of the MLExchange platform, supporting interactive structural exploration across archived datasets and live experiments.

AI 总结：提出物理一致潜在对偶框架，通过联合学习编码器、解码器和潜在流映射，使潜在动力学满足守恒律、不变性和耗散结构，在保持代理模型预测精度的同时提高物理约束满足度和长期行为质量。

链接：https://arxiv.org/abs/2606.15053

机构：Emory University（埃默里大学）； Clemson University（克莱姆森大学）

作者：Matthias Chung, Yutong Bu, Deepanshu Verma

英文摘要：Surrogate models are central to scientific machine learning, where they enable fast prediction, simulation, inference, and control for complex physical systems. For time-dependent problems, however, accurate interpolation of training trajectories is not sufficient: reliable surrogates should also respect the conservation laws, invariants, admissibility conditions, and dissipative structures that give those trajectories physical meaning. We introduce Physics-conforming Latent Twins, a framework for learning latent surrogate solution operators whose dynamics satisfy selected physical principles by design. The method builds on the Latent Twin formulation by jointly learning an encoder, a decoder, and a latent flow map between arbitrary time-indexed states, while constraining the latent dynamics to preserve or dissipate prescribed structural quantities. We develop a constraint-transfer viewpoint that connects physical structure in the original state space with compatible constraints in latent space, and prove structure-preservation bounds showing how latent enforcement improves control of physical defects after decoding. We also derive algebraic conditions for latent flow maps that preserve linear and quadratic invariants or enforce dissipative inequalities. Numerical experiments on representative ODE and PDE benchmarks demonstrate improved constraint satisfaction, structural fidelity, and qualitative long-time behavior while maintaining accurate surrogate prediction.

AI 总结：研究机器学习模型能否超越朴素随机游走基准预测月度美元/加元汇率，采用扩展窗口评估和SHAP解释，发现线性回归显著优于随机游走，集成模型表现接近。

链接：https://arxiv.org/abs/2606.15058

作者：Louis Agyekum, Edmund Fosu Agyemang, Obu-Amoah Ampomah, Kofi Acheampong, Emmanuel Boadi, Priscilla Yaa Amakye, Fafa Shalom Tchorly, Enock Adu Bonsu, Eric Nyarko

英文摘要： This study examines whether machine learning (ML) models can outperform the naive random walk benchmark in forecasting the monthly USD/CAD exchange rate. Using daily data from the Bank of Canada spanning January 2017 to May 2026, resampled into 113 monthly observations, five ML models are evaluated: linear regression, random forest, gradient boosting, XGBoost, and AdaBoost. These models are benchmarked against the naive random walk model and exponential smoothing with Holt-Winters seasonality (ETS). All models are evaluated using an expanding-window framework to maintain strict out-of-sample integrity, and forecast-accuracy differences are assessed using the Diebold-Mariano (DM) test. Structural break detection identifies four significant breakpoints in the series, corresponding to the escalation of the US-China trade war in 2018, the COVID-19 economic recovery in 2020, the peak of the Bank of Canada rate-hiking cycle in 2022, and the start of the Bank of Canada rate-cutting cycle in 2024. SHAP, or Shapley Additive Explanations, analysis is applied to interpret the drivers of the best-performing ML model. The results show that the naive random walk model remains a formidable benchmark. Linear regression is the only model that statistically outperforms the naive random walk model, with a DM statistic of 3.0585 and a p value of 0.0071, whereas the ML ensemble models show only marginal differences. Random Forest with an expanding-window framework achieves the lowest MAPE of 1.17 percent among all models except the random walk. SHAP analysis confirms that short-term lags, particularly lag1 and lag2, and recent rolling means dominate predictions, consistent with the near-random-walk behavior of exchange rates.

AI 总结：提出TriAdReview三角对抗审查架构，使用两个独立审查模型和三角判断机制迭代改进生成器输出，在五个基准任务上相比单模型基线提升10.1%，但发现对抗审查在完整性任务上存在结构偏差。

链接：https://arxiv.org/abs/2606.15074

机构：Hunan Chemical Industry Vocational and Technical College（湖南化工职业技术学院）

作者：Zhiqiang Zhou, Junliang Dai, Xu Ling

英文摘要：Large language models (LLMs) are increasingly used for technical document generation, yet single-model outputs often suffer from over-engineering, security blind spots, and incomplete coverage. We propose TriAdReview, a triangular adversarial review architecture that employs two independent reviewer models (engineering and boundary perspectives) and a triangular judging mechanism to iteratively improve a generator model's output. We evaluate TriAdReview across five benchmark tasks - architecture design, code generation, proposal review, security audit, and requirements analysis - using three configurations: single model (baseline), dual model (single review), and triple model (full system). Results across 75 experiments (n=5 per cell) show that the triple model configuration achieves a 10.1% overall improvement over the single model baseline (26.2 vs. 23.8 out of 50; p<0.05, paired t-test), with particularly strong gains on security audit (+27.6%), code generation (+20.8%), and architecture design (+15.6%). A second scorer (mimo-v2.5-pro) confirms the direction with a smaller effect (+2.7%), suggesting moderate inter-rater agreement. However, the system shows a -7.5% degradation on requirements analysis, revealing that adversarial review architectures have a structural bias toward simplification that is counterproductive for completeness-oriented tasks. We analyze this boundary condition through a task-type framework and demonstrate that reviewer prompt adaptation partially mitigates the issue. Our findings provide the first empirical characterization of when multi-model adversarial review helps versus harms, with implications for the design of collaborative AI systems.

AI 总结：综述知识图谱在医学中的应用，涵盖临床决策支持、疾病预测、健康推荐、精准医疗和医学问答，并讨论构建方法、挑战及未来方向。

链接：https://arxiv.org/abs/2606.15155

机构：Amirkabir University of Technology（阿米尔卡比尔理工大学）

作者：Haniye Sherafatmandjoo, Mohammad Akbari, Zahed Rahmati

英文摘要：Knowledge graphs (KGs) have emerged as a promising solution for integrating and reasoning over complex biomedical and clinical data in healthcare. By representing structured relationships among entities such as diseases, drugs, symptoms, and patient records, KGs provide a semantic backbone for decision-making, prediction, recommendation, and personalized care. Recent advances have demonstrated their utility across diverse medical applications--including clinical decision support systems, disease and treatment outcome prediction, health recommender systems, precision medicine, and medical question answering--where KGs often enhance interpretability, semantic coherence, and patient-specific reasoning. In parallel, a growing body of work focuses on medical KG generation itself, proposing frameworks that construct graphs from EHRs, clinical narratives, biomedical literature, and web resources using ontologies, semantic web technologies, deep-learning-based information extraction, and hybrid neuro-symbolic pipelines. Despite this progress, significant challenges remain, including limited and fragmented knowledge coverage, difficulties in aligning heterogeneous data sources, the fragility of current reasoning and representation-learning methods on dense multi-relational graphs, and unresolved issues related to privacy, bias, and accountability. This survey reviews and categorizes current research on KGs in medicine along both application-oriented and methodology-oriented dimensions, discusses their benefits and technical foundations, and outlines key limitations and open research directions. By analyzing trends, architectures, and evaluation practices, this work aims to guide future developments in KG-driven medical AI systems and support their safe and effective integration into healthcare environments.

AI 总结：提出Edu-Theater框架，通过构建群体水平能力先验和少量诊断查询，利用LLM智能体模拟学习者行为，在减少数据需求的同时提高模拟精度，并增强下游自适应测试等应用。

链接：https://arxiv.org/abs/2606.15225

机构：University of Science and Technology of China（中国科学技术大学）； State Key Laboratory of Cognitive Intelligence（认知智能国家重点实验室）； Southeast University（东南大学）； Alibaba Group（阿里巴巴集团）； iFLYTEK Co., Ltd.（科大讯飞股份有限公司）

作者：Weibo Gao, Qi Liu, Linan Yue, Zheng Zhang, Yichao Du, Fangzhou Yao, Ao Yu, Zhenya Huang, Shijin Wang

英文摘要： Large-scale learner-task interaction data are crucial for intelligent educational systems but are costly to collect and constrained by privacy and learner engagement. Learner simulators play a critical role in simulating scalable learner behavior without the need for continuous involvement of real learners. However, existing methods are predominantly \textbf{individual-centric}, pairing a simulator with each learner to iteratively infer latent knowledge states from dense interaction histories, which is both data- and computation-intensive, and fragile in cold-start scenarios. We propose a \textbf{cohort-aware roll-call simulation paradigm} that first constructs cohort-level proficiency priors and refines individual learner states through a small number of targeted diagnostic queries. Based on this paradigm, we introduce \textbf{Edu-Theater}, an LLM-powered agent system that performs cohort-aware learner simulation via a teacher agent and retrospective roll-call probing over learner logs. Edu-Theater enables scalable future behavior simulation without the need for dense per-learner histories. Experiments on two real-world datasets demonstrate that Edu-Theater achieves higher simulation accuracy with significantly fewer LLM calls, producing synthetic data that enhances downstream applications such as adaptive testing.

AI 总结：提出不确定性感知的贝叶斯深度学习框架，估计2010-2020年伦敦空气污染法规对PM2.5的因果效应，发现法规平均降低PM2.5 1.88 μg/m³（12.35%）。

链接：https://arxiv.org/abs/2606.15257

作者：Yang Han, Jacqueline CK Lam, Victor OK Li, Yiu-Wai Man

英文摘要：Air pollution regulation is central to urban public health governance, but estimating its effects is difficult because policies are implemented non-randomly and pollution trajectories are shaped by meteorology, socioeconomic change, temporal trends, and overlapping interventions. This study develops an uncertainty-aware Bayesian deep learning framework to estimate the aggregate effect of air pollution regulations on PM$_{2.5}$ concentrations in London from 2010 to 2020. The framework integrates daily PM$_{2.5}$ observations from Inner London monitoring stations, meteorological covariates, annual socioeconomic indicators, month-of-year and day-of-week indicators, and daily regulation status data for 32 policy measures. A Bayesian LSTM captures temporal dependencies in environmental and socioeconomic covariates, Bayesian embedding layers represent temporal and regulation status inputs, and a regulation status prediction branch supports propensity score-based adjustment for non-random policy implementation. Regulatory effects are estimated by comparing observed PM$_{2.5}$ concentrations with counterfactual predictions under a hypothetical no-regulation scenario, with uncertainty summarized across repeated Bayesian training runs and bootstrap resampling. Results show that London's regulations were associated with an average PM$_{2.5}$ reduction of 1.88 $\mu$g/m$^3$, a relative reduction of 12.35%, with a 95% confidence interval of 1.64-2.12 $\mu$g/m$^3$. Estimated effects were limited before 2013, became clearer from 2013 to 2017, and were strongest in 2018 and 2019. The findings suggest that sustained and cumulative regulatory interventions contributed to measurable improvements in London's air quality. This study demonstrates how uncertainty-aware causal AI can support environmental accountability, public health protection, and evidence-based governance for environmental decision-making.

AI 总结：提出混合NARX-LLM框架，结合非线性自回归模型与大型语言模型进行残差校正，并引入物理信息提示方法，用于建模格陵兰冰山排放的复杂非线性动态，提升预测准确性。

链接：https://arxiv.org/abs/2606.15288

机构：Heriot-Watt University（赫瑞瓦特大学）； StudioYG

作者：Yiquan Gao, Duohui Xu

英文摘要：Greenland iceberg discharge exhibits complex nonlinear dynamics with limited observability, challenging traditional predictive models. We present a Hybrid NARX-LLM framework that combines a nonlinear autoregressive model with exogenous inputs (NARX) and a large language model (LLM) for residual correction. We further propose a Physics-Informed Prompt (PIP) method that transforms unstructured physical knowledge into structured prompts for zero-shot in-context reasoning. The primary objective is to explore the corrective potential of this framework for modeling Greenland iceberg discharge, rather than merely optimizing predictive accuracy. The NARX component captures intrinsic temporal dependencies, while the LLM, guided by PIP, encodes glacier dynamics and environmental drivers and perceives key trend patterns to correct systematic prediction errors. This integration allows the model to reason about unmodeled factors and produce interpretable residuals, enhancing overall predictive accuracy. Applied to Greenland iceberg discharge time series, our approach addresses extreme events that are difficult to predict due to rare variations and nonstationary trends, a limitation often overlooked by traditional methods. By fusing structured time-series modeling with knowledge-driven foundation AI, the framework offers a scalable and interpretable pathway to bridge data-limited climate forecasting with physics-informed LLM reasoning. The code is available.

AI 总结：研究在工业表格数据中，LLM（嵌入、直接分类、混合堆叠）与经典树集成方法的对比，发现LLM在语义受限时效果有限，但嵌入和混合方法仍有价值。

链接：https://arxiv.org/abs/2606.15314

机构：Technical University of Munich（慕尼黑工业大学）； BMW Group（宝马集团）

作者：Aina Vila Pons, Ioannis Tzachristas, Constantinos Antoniou

英文摘要： Industrial retrofit planning depends on structured operational data rather than free text: planners must estimate whether a newly registered prototype will require a retrofit, which retrofit package it will need, and how long the work will take. We study an industrial dataset linking a prototype-registration system (284,271 vehicles) with a retrofit-management system (48,716 cleaned visits), and compare strong tabular machine learning baselines with three LLM-based strategies on row-serialized inputs: embedding features (Amazon Titan), direct prompted classification (Claude Sonnet 4), and an ML+LLM stacking approach. Across binary occurrence prediction, 15-way retrofit-type classification, per-visit duration regression, and an aggregated monthly benchmark, classical tree ensembles remain the strongest standalone models. However, the LLM results reveal a consistent pattern: embeddings remain useful on tables (binary AUC = 0.982), direct prompting collapses once semantic signal is stripped by hashing (binary AUC = 0.500; multiclass weighted F1 = 0.018), and hybrid stacking yields the best manually built multiclass model (weighted F1 = 0.626). On the monthly benchmark, lag-based machine learning outperforms time-series foundation models, though Chronos-small remains competitive in zero-shot forecasting. The results suggest that on privacy-constrained industrial tables, LLMs are more effective as complementary components than as replacements for strong tabular baselines.

AI 总结：提出标签噪声对比鲁棒学习（LaNCoR）方法，通过对齐波形特征与标签表示分布来纠正错误标签，在微地震P波到时拾取任务中性能提升高达28.8%。

链接：https://arxiv.org/abs/2606.15377

作者：Sen Li, Xu Yang, S. Mostafa Mousavi, Anye Cao, Keting Fan, Yaoqi Liu, Changbin Wang, Qiang Niu

英文摘要：Inaccurately labeled training data, or "label noise", poses a significant threat to the integrity of supervised machine learning models. This corruption directly degrades performance by teaching the model erroneous mappings between features and labels, which leads to poor generalization and reduced accuracy on properly labeled validation and test data. Current seismological applications mainly rely on large-scale training sets or data augmentation to reduce the label-noise impact, which can be labor-intensive and costly. Here, we introduce a Label Noise-Contrastive Robust Learning (LaNCoR) approach that can effectively handle noisy labels in seismic signal processing tasks, without requiring large-scale training datasets. In this approach, the input waveform feature and label representation distributions are aligned in the feature space to correct mislabeling and reduce its impact on the training process. We present LaNCoR's performance on the task of P-phase arrival-time picking of real microseismic data using two baseline models and training approaches. Our results indicate that LaNCoR can improve performance by up to 28.8% across performance metrics. This approach holds great promise for model training in seismology and geosciences.

AI 总结：研究利用提示驱动的视觉语言模型在轨扩展语义能力，无需修改权重即可通过自然语言提示检测新航天器部件，在129张图像上零样本实例分割达到0.385 mAP@0.5。

链接：https://arxiv.org/abs/2606.15427

机构：Florida Institute of Technology（佛罗里达理工学院）； University of Florida（佛罗里达大学）

作者：Nicholas A. Welsh, Lennon J. Shikhman, Monty Nehru Attazs, Seemanthini K. Putane, Van Minh Nguyen, Ryan T. White

英文摘要：Spaceborne inspection systems often deploy perception models prior to launch, after which updating model weights or expanding fixed label sets becomes operationally impractical. While supervised models can be integrated pre-flight, adding new semantic capabilities in orbit requires retraining and re-uploading parameters. We investigate whether prompt-driven vision--language models can enable post-launch semantic expansion, allowing new spacecraft components to be specified via natural-language prompts without modifying onboard weights. We evaluate zero-shot instance segmentation of spacecraft components under a strictly frozen, single-pass inference protocol on a test set of $129$ images of previously unseen satellites. Under fixed global thresholds and no post-processing, SAM3 achieves $0.385$ mAP@$0.5$ and $0.267$ mAP@$0.5{:}0.95$. Performance is strongly scale-dependent: large structural elements like spacecraft bodies ($0.639$ AP@$0.50$) and solar arrays ($0.598$ AP@$0.5$) localize reliably, while relatively small appendages like antennas ($0.221$ AP@$0.5$) and thrusters ($0.081$ AP@$0.5$) remain difficult. Prompt formulation influences performance, with structured prompts incorporating spatial and geometric descriptors yielding up to $82%$ improvement over short category-name prompts. The model operates within the memory and compute envelope of contemporary embedded GPUs, suggesting prompt-driven grounding can provide a practical mechanism for post-launch semantic extension of dominant spacecraft structures while highlighting limitations of zero-shot localization for fine-scale components under orbital domain shift.

AI 总结：提出惊喜引导的归并排序（SGS）框架，利用视觉语言模型（VLM）作为问题优先级排序器，通过自适应预算分配将高模糊度比较路由给人类，在六个基准上以相同预算实现Kendall's τ×100提升6-12点。

链接：https://arxiv.org/abs/2606.15623

机构：Hanyang University（汉阳大学）； Hankuk University of Foreign Studies（韩国外国语大学）

作者：Yujin Park, Haejun Chung, Ikbeom Jang

英文摘要：Pairwise comparison is the gold standard for subjective ranking tasks; however, exhaustive annotation requires a massive number of human comparisons ($O(n^2)$). While sorting-based methods have reduced this burden to $O(n\log n)$, they still require expensive human judgment for every single comparison. To further improve annotation efficiency, we propose leveraging a Vision-Language Model (VLM) not as an annotator replacement, but as a \emph{question prioritizer} to identify which comparisons genuinely require human judgment. The proposed \textbf{Surprise-Guided MergeSort (SGS)} framework achieves this through three integrated components: (1) a bottom-up MergeSort scheduler that structures comparisons and exploits transitivity, (2) a composite Surprise Scorer -- combining position-bias-cancelled VLM confidence, Elo gap, and vote entropy -- to quantify comparison ambiguity, and (3) an adaptive budget allocator that routes high-surprise pairs to humans while automating low-surprise pairs via transitivity inference. Validation was conducted on six diverse benchmarks spanning text similarity (STS-B, BIOSSES, SICKR-STS) and image quality assessment (KonIQ-10k, TID2013, LIVE Challenge). SGS effectively identified and skipped up to 535 non-informative comparisons per session. Consequently, it achieved Kendall's $\tau{\times}100$ improvements of $+6$ to $+12$ over Active Elo under the same total budget. These results demonstrate that combining VLM-guided surprise metrics with algorithmic sorting provides a generally consistent accuracy-efficiency trade-off across diverse domains.

AI 总结：提出HAPI框架，通过物理集成灰盒模型、元学习快速自适应和条件生成模型，构建可识别、强预测性的心脏电生理数字孪生。

链接：https://arxiv.org/abs/2606.15637

机构：Rochester Institute of Technology（罗切斯特理工学院）

作者：Sumeet Vadhavkar, Xiajun Jiang, Yubo Ye, Maryam Toloubidokhti, Linwei Wang

英文摘要：A digital twin (DT) of a patient-specific heart offers significant potential in personalized medicine. However, its rapid and dynamic adaptation to an individual's live data and its predictive capability after adaptation remains central challenges. We examine this challenge from its two building blocks: DT formulation where mechanistic and data-driven models show competing merits and limitations, and DT optimization strategies that are largely driven by a reconstruction objective leading to un-identifiable models. We address both bottlenecks via HAPI -- an AI framework for building hybrid, adaptive, and predictive DTs with three key enablers. First, HAPI constructs a physics-integrated gray-box model in which an interpretable mechanistic backbone is augmented by a neural component that models its residual to the observed data. Second, rather than attempting to pre-encode all possible variations in a static hybrid model, HAPI enables rapid on-the-fly adaptation of the hybrid model to few-shot live data, achieved by feedforward meta-learners realizing amortized inference of both mechanistic and neural parameters of the hybrid model trained with predictive objectives. Finally, we show that this adaptivity corresponds to the construction of a conditional generative model (i.e., the hybrid DT) that endows it with theoretical identifiability and thus strong performance in predictive scenarios. We demonstrate the proof-of-concept of HAPI in cardiac electrophysiology using a hybrid monodomain model with mechanistic reaction kinetics and neural graph diffusion. Across synthetic and real-data studies, we show that HAPI's mechanistic-neural hybridization and predictive adaptation are critical for obtaining identifiable DTs with strong predictive and out-of-distribution capabilities.

AI 总结：提出UMAR框架，通过三个专业智能体独立评估风险并校准不确定性，利用Dempster-Shafer理论融合分数并测量冲突，在SEC 10-K数据集上优于基线模型，提供可解释的风险信号。

链接：https://arxiv.org/abs/2606.15640

作者：Yuhan Wang, Manqing Wang, Yixuan Lu, Zhaoyue Peng, Shengda Lin

英文摘要：Audit risk assessment increasingly benefits from combining heterogeneous evidence sources, yet existing approaches typically produce point predictions without quantifying how well different evidence streams agree. We propose UMAR (Uncertainty-Aware Multi-Agent Risk Assessment), a framework that employs three specialized agents: an MD&A Text Agent, a Financial Ratio Agent, and a CAM Agent, each producing independent risk scores with calibrated uncertainty estimates. An Uncertainty Aggregator based on Dempster-Shafer evidence theory fuses these scores while explicitly measuring inter-agent conflict. We evaluate UMAR on a U.S. dataset of 3,200 firm-year observations from SEC 10-K filings (2019-2023), with financial restatement as the target label. Experimental results show that UMAR achieves an AUROC of 0.782 and a PR-AUC of 0.341, outperforming logistic regression, XGBoost, FinBERT, and single-agent and dual-agent LLM baselines. UMAR attains the lowest expected calibration error (ECE = 0.052) among all methods and identifies evidence-conflict patterns that correlate with actual restatement risk, offering auditors potentially actionable and interpretable risk signals.

AI 总结：针对跨城市数据稀缺场景，提出CIWI-CKT框架，融合混沌信息波生成、元干涉处理和混沌感知元学习，显著提升预测精度并降低数据需求。

链接：https://arxiv.org/abs/2606.15642

机构：Griffith University（格里菲斯大学）； School of Information and Communication Technology, Griffith University（格里菲斯大学信息与通信技术学院）； School of Information Engineering, Chang’an University（长安大学信息工程学院）

作者：Abdul Joseph Fofanah, Lian Wen, David Chen, Shaoyang Zhang

英文摘要：Accurate traffic flow prediction remains challenging in cross-city, data-scarce scenarios where limited historical data hinders model generalisation. The chaotic nature of traffic dynamics, complex spatio-temporal dependencies, and heterogeneous urban networks complicate few-shot learning across cities. Existing deep learning approaches either treat traffic as purely deterministic or lack mechanisms to model wave-like interference patterns essential for cross-regime traffic dynamics. To address these limitations, this paper proposes CIWI-CKT, a novel Chaos-Informed Wave Interference Feature Fusion framework with Cross-City Knowledge Transfer. Our framework introduces three core innovations: chaos-informed wave generation that extracts measurable chaos invariants and models traffic as adaptive wave components; meta-interference processing that captures wave interactions between support and query regimes while producing a predictability score for confidence estimation; and chaos-aware meta-learning that enables efficient cross-city knowledge transfer while preserving chaotic characteristics. We establish theoretical guarantees including chaos-to-wave stability, wave-induced dimension reduction, and meta-learning generalisation bounds. Extensive experiments on four real-world traffic datasets demonstrate that CIWI-CKT significantly outperforms state-of-the-art spatio-temporal graph learning, transfer learning, prompt-based, and few-shot methods, improving prediction accuracy while substantially reducing required training data.

AI 总结：提出改进的Transformer架构结合余弦退火学习率调度和移位数据增强（SDA），在VN30和S&P 500指数上有效降低预测误差和波动性，优于增加模型复杂度的方法。

链接：https://arxiv.org/abs/2606.15701

机构：Faculty of Mathematics and Statistics, Ton Duc Thang University（孙德胜大学数学与统计学院）

作者：Tien Thanh Thach

英文摘要：Transformers have shown remarkable success in sequence modeling, yet their direct application to financial time series remains challenging due to noisy signals, short-memory dynamics, and distributional shifts. This paper proposes a modified Transformer architecture for one-step stock index forecasting, combined with advanced learning-rate scheduling and a novel Shifted Data Augmentation (SDA) technique. We evaluate the proposed framework on two benchmark stock index datasets, VN30 and S&P 500. Experimental results demonstrate that cosine annealing with warmup consistently improves forecasting accuracy over the generalized inverse-power scheduler. Furthermore, SDA substantially reduces forecasting errors and run-to-run variability while improving robustness to hyperparameter selection. The combination of cosine annealing scheduling and SDA achieved the best performance on both datasets, indicating that data augmentation can play a more important role than increasing model complexity in Transformer-based financial forecasting. These findings provide a practical and computationally efficient approach for robust stock index forecasting in noisy financial environments.

AI 总结：提出基于因果推断的换道预测框架，结合深度结构因果建模与干预效应分析，在预测准确率超过95%的同时，识别直接贡献变量及其因果链，实现可解释的因果推理。

链接：https://arxiv.org/abs/2606.15756

机构：University of Alcalá（阿尔卡拉大学）

作者：Mohamed Manzour, Aditya Kumar, Augusto Luis Ballardini, Miguel Ángel Sotelo

英文摘要：Lane-change prediction is a central task in intelligent vehicles, where early maneuver anticipation can support safer decision-making. However, many existing approaches mainly learn statistical associations between observed driving variables and future maneuvers, while overlooking the causal dependencies among the input variables themselves. This limits interpretability, especially when physically related variables such as longitudinal gap, relative longitudinal velocity, and Time-To-Collision (TTC) are treated as independent flat inputs. This article presents a causal-inference-based framework for lane-change prediction and explanation. The proposed approach combines linguistic feature construction, expert-constrained causal discovery, deep structural causal modeling with Deep End-to-end Causal Inference (DECI), intervention-based effect analysis, refutation testing, and recursive causal-chain explanation. The objective is not only to predict the future maneuver, but also to identify candidate variables that directly contribute to the prediction, the upstream factors influencing them, and the causal chains through which these effects propagate. The framework achieves average F1-scores above 95% during the first three seconds before the lane-marking crossing event. Beyond prediction accuracy, the framework uses intervention-based effect analysis to distinguish influential from weakly influential variables under the learned causal structure. It further distinguishes candidate direct contributors from mediated effects and generates contrastive causal-chain explanations that clarify why the predicted maneuver is favored and why the alternative maneuvers are less supported. The main contribution is therefore a mechanism-aware lane-change prediction pipeline that moves beyond correlation-based classification toward more interpretable causal reasoning for maneuver prediction.

AI 总结：提出BN-LTE框架，结合贝叶斯网络与潜在时间嵌入，利用AT(N)级联约束建模AD进展，在ADNI数据上优于基线，并识别出淀粉样蛋白敏感性的中期伪时间窗口。

链接：https://arxiv.org/abs/2606.15784

机构：Alzheimer's Disease Neuroimaging Initiative（阿尔茨海默病神经影像学倡议）； Open Access Series of Imaging Studies（开放获取影像学研究系列）

作者：Nguyen Linh Dan Le

英文摘要：Alzheimer's disease (AD) progression is often described through the amyloid-tau-neurodegeneration, or AT(N), cascade. However, most longitudinal models represent this cascade either as a fixed sequence of biomarkers or as a black-box forecasting task. This makes it difficult to determine when biologically guided biomarker relationships influence future regional pathology. In this study, we introduce Bayesian Networks with Latent Time Embedding (BN-LTE), a Bayesian structural framework for stage-aware modeling of AD progression. BN-LTE estimates disease pseudotime from baseline biomarker profiles and constrains directed dependencies according to biologically plausible AT(N) ordering. Posterior spline-varying structural equations are then used to link initial multimodal measurements with future annualized regional tau-PET change. Across repeated subject-disjoint evaluations using ADNI data, BN-LTE shows strong spatial reconstruction of tau progression compared with the included forecasting baselines. Beyond spatial reconstruction, BN-LTE recovers posterior stage-varying AT(N)-constrained effects and identifies a mid-pseudotime window of amyloid sensitivity. This window is supported by model-implied g-formula contrasts, root-adjusted AIPW, mechanism-sensitive ablations, and robustness analyses across spline and prior specifications. Overall, these findings position BN-LTE as a Bayesian structural framework for forecasting tau progression while examining stage-dependent AT(N)-cascade mechanisms in observational longitudinal neuroimaging data. Our code is available at this https URL.

AI 总结：提出记忆增强图液态时间常数网络（MA-GLTC），通过时空单元分解、图液态时间常数动态和记忆迁移存储机制，实现连续时间下的跨域交通状态预测，在五个数据集上优于现有方法。

链接：https://arxiv.org/abs/2606.15807

机构：Software College, Liaoning Technical University（辽宁工程技术大学软件学院）

作者：Jinrong Xiang, Ming Xu

英文摘要： Traffic state prediction is a fundamental task in intelligent transportation systems. In practical applications, some regions suffer from limited traffic observations due to insufficient sensing infrastructure, making cross-domain knowledge transfer an important solution for data-scarce traffic prediction. However, existing cross-domain traffic prediction methods still face several limitations, including coarse-grained source-target adaptation, limited capability in handling unseen target-domain patterns, and insufficient modeling of continuous traffic dynamics under irregular or heterogeneous temporal conditions. To address these issues, this paper proposes a continuous cross-domain traffic prediction framework, termed Memory-Augmented Graph Liquid Time-Constant Network (MA-GLTC). Specifically, we first construct spatio-temporal units (STUs) to decompose traffic networks into transferable local units, enabling fine-grained knowledge alignment across domains. Then, a graph liquid time-constant network (GLTC) is developed to model graph-coupled traffic evolution in continuous time. Different from generic graph neural ODE-based models, GLTC introduces graph-coupled recurrent conductance into liquid time-constant dynamics, allowing node states to evolve with leakage, adaptive time constants, and neighborhood-aware feedback. Furthermore, a Memory-based Transfer Storage (MTS) mechanism is designed to preserve source-domain knowledge, retrieve matched traffic patterns, and update reliable target-domain patterns when unseen states emerge. Experiments on five public traffic datasets demonstrate that MA-GLTC consistently outperforms representative innerdomain and cross-domain baselines in both short-term and longterm prediction tasks. Compared with the second-best method, MA-GLTC reduces the average prediction errors by 3.02%, 0.33%, 8.92%, 10.09%, and 2.11%, respectively.

AI 总结：提出标量路径修正方法（PAN池化和PGS混合器），在保持等变骨架不变下优化标量通道，使MACE等势的力误差降低22-27%，能量误差降低19-22%，且计算开销仅增5%。

链接：https://arxiv.org/abs/2606.15892

机构：Science and Technology Facilities Council（科学技术设施委员会）； Diamond Light Source（钻石光源）

作者：Jia Bi, Alin Marin Elena, Samuel Pinilla

英文摘要：Accurate interatomic potentials enable molecular dynamics of materials, molecules, and interfaces beyond density-functional-theory length and time scales. Equivariant neural network potentials have improved the representation of local geometry. However, their deployable energy surfaces ultimately manifest through invariant scalar channels, whose aggregation and spectral resolution remain comparatively underexamined. Here we use Physics-Aware Neighborhood (PAN) pooling and Physics-Guided Spectral (PGS) mixers as controlled scalar-pathway probes: lightweight, symmetry-preserving modifications that act only on \(\ell=0\) channels while leaving the equivariant tensor backbone unchanged. Using MACE as a high-body-order mechanistic scaffold, PAN adds coordination-sensitive amplitude modulation, whereas PGS augments edge and readout scalar features with radial and tapered spectral bases. Across metallic Ag, covalent Si, a short-range ionic LiF/Li--F subset, and MD17/rMD17 molecules, this scalar-pathway correction reduces MACE force errors by 22--27\% and energy errors by 19--22\%; on systems with stress labels, stress errors decrease by 27--28\%, at approximately 5\% additional inference-FLOPs cost. Directionally consistent gains in Allegro and NequIP further indicate that the correction is portable across distinct short-range equivariant backbones, although effect sizes remain architecture-dependent. These results identify scalar-pathway fidelity as a practical design dimension for short-range equivariant interatomic potentials.

AI 总结：本研究利用智能手表PPG信号和CGM血糖数据构建机器学习模型，探索无创血糖估计的可行性，初步结果显示存在预测信号但需更多数据验证。

链接：https://arxiv.org/abs/2606.15927

机构：Indraprastha Institute of Information Technology, Delhi（德里印度信息技术学院）

作者：Ruhani Bhatia, Vijval Ekbote

英文摘要：Diabetes and extreme blood sugar levels are some of the major health problems faced by humans today across the world. While Continuous Glucose Monitoring (CGM) has emerged as an effective technology for management of diabetes as well as for monitoring blood sugar levels, this technology has traditionally been invasive (that is, requiring the piercing of the skin) and carries the risk of irritation, induration, etc. This highlights the need for accurate and non-invasive CGM methods that can be deployed at scale. With the emergence of various sensing technologies and their integration in wearables like the smart-watch, we now have the capability to continuously monitor body signals like the Photoplethysmogram (PPG) in a non-invasive manner. Having the ability to continuously monitor blood glucose through CGMs and continuously monitor PPG signals through a smart-watch offers an opportunity to get dense data on these two, opening the possibility of building machine learning and deep learning based models to estimate blood glucose level from PPG signals. In this work, we first present a paired dataset comprising continuous PPG signals from a smartwatch along with glucose values recorded using a CGM device. We also present the results of some preliminary experimental explorations performed on our dataset. These preliminary results suggest that some predictive signals may exist, though more exploration is needed with more data from a larger number of individuals. The dataset can be accessed at this https URL

AI 总结：提出IBAD框架，利用LDA学习可解释的日常移动模板，通过层次自监督模型检测个体行为异常，在真实和合成数据集上验证了模板的可迁移性和鲁棒性。

链接：https://arxiv.org/abs/2606.16023

作者：Bita Azarijoo, John Krumm, Cyrus Shahabi

英文摘要： Human mobility appears highly diverse, yet much of a person's daily mobility can be explained by a small set of recurring behavioral templates, such as commuting, school-centered activities, caregiving, nightlife, or errand patterns. We present \texttt{IBAD} (\underline{I}nterpretable \underline{B}ehavioral \underline{A}nomaly \underline{D}etection), a framework that learns interpretable daily mobility templates and represents each individual as a distribution over mixtures of these templates. Rather than focusing on specific locations, IBAD characterizes activities that individuals perform across locations. This approach first discovers global behavioral templates using Latent Dirichlet Allocation (LDA), then employs a hierarchical self-supervised model to learn normal behavior of individuals from their soft behavioral templates. We also introduce a \emph{splicing benchmark} that creates controlled behavioral mismatches between an individual's historical profile and injected mobility patterns. Experiments on real-world and synthetic datasets show that daily behavior can be effectively decomposed into a small number of interpretable templates. Crucially, we show that the learned behavioral archetypes \emph{transfer} across distinct geographic and demographic contexts. Furthermore, IBAD maintains a robust competitive performance across all settings. For reproducibility purposes, the code is accessible at ~\href{ this https URL }{ this https URL }.

AI 总结：利用NHANES数据训练LightGBM等六种机器学习模型，实现无需实验室检测的血糖异常风险筛查，AUC达0.820，优于传统风险评分，并识别出年龄、种族和腰高比等关键预测因素。

链接：https://arxiv.org/abs/2606.16056

机构：Department of Computer Science, Aarhus University（奥胡斯大学计算机科学系）； University at Buffalo, SUNY（纽约州立大学布法罗分校）

作者：Black Sun, Chenyi Zhang, Kaiyi Ji, Xi Lu

英文摘要：Dysglycemia, encompassing both prediabetes and diabetes, affects huge numbers of adults worldwide, yet many of them remain undiagnosed. We developed and validated machine-learning (ML) models for non-invasive screening of dysglycemia risk that require no laboratory tests. Pooling data from the National Health and Nutrition Examination Survey (NHANES) 2017--2023 (n=14,352), we trained six ML models with stratified 5-fold cross-validation and compared them with two established clinical risk scores. LightGBM achieved the highest area under the receiver operating characteristic curve (AUC=0.820, 95% CI: 0.806--0.835), outperforming the Finnish Diabetes Risk Score (0.745) and American Diabetes Association Risk Test (0.783). SHAP analysis identified age, race/ethnicity, and waist-to-height ratio as the most influential predictors. Subgroup analyses confirmed consistent performance across demographic strata (AUC: 0.735--0.832). These results demonstrate the feasibility of explainable, laboratory-free dysglycemia screening for deployment in community settings and self-tracking health applications.

AI 总结：本研究系统评估EEGNet在fNIRS认知负荷分类中的性能，发现重叠分段和小固定学习率在随机分割中表现最佳，但受试者独立评估准确率大幅下降，非重叠分段和PCA特征在SI评估中取得最佳56.11%准确率，表明消除时间冗余有助于学习更鲁棒的跨个体表征。

链接：https://arxiv.org/abs/2606.16160

作者：Mehshan Ahmed Khan, Houshyar Asadi, Li Zhang, Mohammad reza Chalak Qazani, Ghazal Bargshady, Stefanos gkikas, Christian arzate, Sam Oladazimi, Zoran Najdovsk, Lei Wei, Chee Peng Lim

英文摘要：Accurately classifying cognitive load from functional near-infrared spectroscopy (fNIRS) signals remains a significant challenge due to temporal variability, inter-subject differences, and sensitivity to preprocessing choices. This study provides a comprehensive evaluation of EEGNet for fNIRS-based cognitive load classification by systematically examining the effects of temporal segmentation strategies (overlapping vs. non-overlapping), window lengths (10s, 20s, 30s), feature extraction methods (Analysis of Variance (ANOVA), Principal Component Analysis (PCA), Fast Independent Component Analysis (FastICA)), learning rate configurations (fixed and adaptive), and evaluation protocols (random split vs. subject-independent (SI)). Results from random-split experiments show that overlapping segmentation, combined with smaller fixed learning rates (0.01-0.001), yields the highest accuracies, due to temporal redundancy and dense sampling of hemodynamic transitions. However, SI evaluation reveals a substantial drop in accuracy, demonstrating limited generalization to unseen participants. Under SI evaluation, non-overlapping segmentation outperformed overlapping windows, with the best accuracy of 56.11% achieved using PCA features with a 20-second window and a 0.1 learning rate. These findings indicate that eliminating temporal redundancy helps the model learn more robust and generalizable representations of cognitive load across individuals. Although adaptive learning rate strategy improved training stability, it did not surpass the performance of optimally selected fixed learning rates. The study highlights the critical role of segmentation strategy and learning rate selection in improving model generalization and identifies methodological considerations essential for developing reliable, real-time, and SI cognitive load classification systems using fNIRS.

AI 总结：提出一种LLM驱动的虚拟人群模型，通过混合客户画像和LLM评估购买概率，生成需求分布，支持风险感知定价，在H&M数据集上表现最优。

链接：https://arxiv.org/abs/2606.16183

机构：Columbia University（哥伦比亚大学）

作者：Chengpiao Huang, Kaizheng Wang

英文摘要： We develop an LLM-powered virtual population model that simulates demand for pricing decisions, in settings where products are described by rich unstructured information, such as text descriptions and images, and where decision makers need not only mean-demand predictions but also uncertainty estimates for counterfactual prices. Our model represents exposed customers as draws from a finite mixture of customer personas. For each persona, product, and candidate price, an LLM elicits a persona-level purchase probability using both structured persona information and unstructured product information. These probabilities are aggregated through calibrated mixture weights to form a predictive distribution of aggregate demand. The resulting simulator can evaluate counterfactual prices under various pricing objectives, including expected revenue and risk-aware criteria such as conditional value at risk. We test the framework on an online H&M fashion dataset with product descriptions and images. The calibrated LLM-based simulator achieves the best overall predictive performance among the models considered, and supports sample-efficient pricing decisions. Our framework provides a practical way to use LLMs as demand simulators for products with limited historical demand data but rich product information. By producing a full predictive demand distribution rather than only a point forecast, it enables managers to compare candidate prices, quantify demand uncertainty, and choose prices that target either average-case revenue or risk-aware objectives.

AI 总结：提出基于主动学习和生成学习的机器学习方法，结合梯度提升回归树模型，预测并行化学计算的运行时参数，在CCSD计算中MAPE低至0.023，R²高达99.9%。

链接：https://arxiv.org/abs/2606.16226

作者：Tanzila Tabassum, Omer Subasi, Ajay Panyala, Epiya Ebiapia, Gerald Baumgartner, Erdal Mutlu, P Sadayappan, Karol Kowalski

英文摘要：In this work, we develop two main Machine Learning based approaches to predict the runtime parameters of highly scalable parallel chemistry this http URL approaches employ active and generative learning together with the empirically determined gradient boosted regression tree models chosen among a rich suite of machine learning models. When evaluated on Coupled-Cluster with Singles and Doubles computations, our models achieve a mean absolute error percentage (MAPE) as low as 0.023 and a coefficient of determination as high as 99.9%. Furthermore, when combined with active learning to mitigate the lack of large amounts of training data, our models score a MAPE about 0.2 with 20-25% of the original dataset.

AI 总结：提出TC-SOH模块化服务架构，通过时间对比机制和跨窗口预测任务从原始数据中提取退化相关表示，实现自主端到端SOH预测，在四个数据集上MAPE和RMSE分别降低1.91倍和2.13倍。

链接：https://arxiv.org/abs/2606.16434

作者：Junting Wen, Dan Li, Qihao Quan, Xiwen Wang, Hang Yang, Zhaohong Meng, Zigui Jiang, Changlin Yang, Tianle Liu, Diego Muñoz-Carpintero, Jian Lou

英文摘要：Accurate state of health (SOH) estimation is a critical diagnostic service for lithium-ion battery management. However, reliance on labor-intensive manual feature engineering and opaque black-box models hinders scalable industrial deployment. To address this, we introduce TC-SOH: a modular, plug-and-play service architecture for autonomous, end-to-end SOH prediction. TC-SOH employs a temporal-contrastive mechanism and a cross-window prediction pretext task to extract degradation-relevant representations directly from raw operational data. To improve transparency, we connect model efficacy with representation diagnostics: visualization, sensitivity analysis, redundancy analysis, bidirectional probing, future-SOH probing, and temporal shuffling show that learned features overlap with selected expert descriptors while retaining additional SOH-relevant variation, and that ordered temporal context improves subsequent-SOH prediction. Across four public datasets, TC-SOH outperforms the considered physics-informed and data-driven baselines, reducing MAPE by 1.91 times and RMSE by 2.13 times.

AI 总结：提出SpTGNN，一种多模态时空图神经网络，通过异构图注意力、微调基础模型特征提取和稀疏专家混合融合，结合异方差回归与深度集成的不确定性量化，在三个区域数据集上优于XGBoost基线。

链接：https://arxiv.org/abs/2606.16580

机构：Spatialise B.V.

作者：Daniele Mos, Felipe Drummond, Anton Bossenbroek, Soufiane el Khinifri

英文摘要：Top-soil organic carbon (SOC) prediction is fundamental to agricultural sustainability, land use policy and fertilization planning. Existing approaches face two limitations: they pair hand-crafted covariates with classical ML or single-modal deep models that miss rich spectral and temporal information, and grid-based architectures ignore the irregular spatial structure of field measurements. We introduce SpTGNN, a multi-modal spatio-temporal graph neural network addressing both. SpTGNN represents soil measurements as nodes in a heterogeneous graph with three edge types (spatial proximity, spectral similarity, elevation), and applies relational graph attention to learn separate patterns per relation. A fine-tuned TerraMind encoder extracts node features from Sentinel-2, Sentinel-1 and DEM signals, combined with per-sample environmental covariates and learned positional and temporal embeddings. A sparse Mixture-of-Experts module fuses the four streams via top-$k$ routing. Uncertainty is captured by pairing heteroscedastic regression (aleatoric) with deep ensembles (epistemic), and a Moran's $I$ penalty regularizes spatial autocorrelation. We evaluate on a global SOC corpus split into three regional instances ($\sim$49k samples globally, Africa $\sim$26k, Europe $\sim$14k). Our 5-member deep ensemble reports $R^2=0.762$, RMSE $=3.51\pm0.48$ g/kg and MAPE $=22.9\%$ on the Africa test split, improving over a tabular XGBoost baseline; the best single checkpoint reaches validation $R^2=0.864$. Ablations confirm the heterogeneous graph, MoE fusion and fine-tuned backbone each contribute substantively, and the ensemble UQ stack achieves post-calibration ECE of $0.031$ (hybrid) and $0.026$ ($\beta$-NLL). To our knowledge, this is the first framework to unify foundation-model feature extraction, heterogeneous graph attention and decomposed uncertainty quantification for SOC estimation.

AI 总结：本文利用挪威保险公司的生产数据，训练梯度提升决策树模型检测洗钱索赔，并引入欺诈标签辅助训练，在预算加权捕获率指标下，最佳模型在2-6%的审查索赔中捕获近三分之二的洗钱案例。

链接：https://arxiv.org/abs/2606.16663

作者： Dara Goldar, Geir Kjetil Ferkingstad Sandve, Martin Jullum

英文摘要：Money laundering through insurance claims poses a threat to insurers both through fraudulent payouts and reputational and regulatory risk. Despite this, little research has examined how such laundering can be prevented. This paper examines whether machine learning can help insurers flag suspicious claims before payout, shifting the focus from passive reporting to active prevention. Using production data from a major Norwegian insurer, we train gradient-boosted decision tree models to detect claims later reported to authorities for suspected money laundering. Because fraud and laundering may share behavioural patterns, we also examine whether insurance fraud labels can serve as an auxiliary training signal. We compare different learning setups using the Budget-Weighted Capture Rate, a metric introduced in this paper to measure how many laundering cases are captured when only a small share of claims can be manually reviewed. The results show that incorporating fraud-related investigation labels substantially improves laundering detection. The best-performing model captures nearly two-thirds of laundering cases within the top-ranked 2 to 6 percent of claims selected for investigation. To our knowledge, this is the first empirical study of machine learning for money laundering detection in insurance claims.

AI 总结：提出集成营销归因（IMA）框架，结合营销组合模型（MMM）与贝叶斯归因模型，从聚合数据中推导出活动级效果，实现隐私安全且粒度精细的归因。

链接：https://arxiv.org/abs/2606.16878

作者：Meghana R. Bhat, Ankit Umare, Utsav Aggarwal, Richard Vecsler, Arunkumar Mani, Karthik Nair, Chandhu Nair

英文摘要：Retail marketing measurement increasingly requires granular campaign-level insights without relying on user-level tracking. However, the two dominant approaches, Marketing Mix Modeling (MMM) and Multi-Touch Attribution (MTA), often produce fragmented insights. MMM is privacy-safe and robust for channel-level planning but is too coarse for campaign optimization, while MTA provides granular attribution but has become less reliable under increasing privacy restrictions. We propose Integrated Marketing Attribution (IMA), a unified framework that combines MMM with channel specific Bayesian attribution models to derive campaign-level effects from aggregated data. By leveraging MMM-informed priors, IMA delivers granular, privacy-safe attribution while preserving consistency with MMM.

AI 总结：提出混合卷积VAE结合二次微笑重拟合的预测器，在BTC和ETH期权数据上实现低RMSE，显著优于纯参数化方法，并消除日历和蝶式套利。

链接：https://arxiv.org/abs/2606.16961

机构：Jasper Research, USA（Jasper Research（美国））

作者：Sadanand Singh, Allam Reddy, Manan Chopra

英文摘要：We present a convolutional variational autoencoder for cryptocurrency implied-volatility surfaces, together with a deployable predictor that combines it with a quadratic smile re-fit through a deterministic per-tenor routing rule. Trained on 6,034 fully-filled hourly Binance Options surfaces of BTC and ETH spanning May-October 2023 and parameterised on a common $6 \times 7$ tenor-delta grid, the model attains a hidden-cell surface-completion RMSE in the 0.94-1.56 vol-point range across both markets and mask rates 10-50%. The hybrid predictor attains 0.83 vol points at 50% masking against 7.00 for the smile re-fit alone, an eightfold reduction obtained at no additional inference cost. Under structurally-correlated hole patterns that emulate the withdrawal of an entire tenor of strikes, the smile re-fit incurs 9.6-13.1 vol points of error while the learned model remains at 1.5-1.9, isolating a regime in which the generative model is the only viable predictor. Joint training on BTC and ETH improves the in-distribution model on both markets by 9-27% relative to the better-performing single-symbol counterpart, indicating a substantially shared vol-surface manifold across the two largest cryptocurrencies over the observation window. The hybrid is calendar- and butterfly-arbitrage-free at the listed strikes, a property that the parametric smile re-fit alone fails at high mask rates. The per-snapshot reconstruction error of the trained model flags the late-October ETF-anticipation rally and the August $17$, $2023$ flash crash as elevated-error periods without supervision. All training and evaluation infrastructure is released to support reproducible follow-on work.

AI 总结：提出NEXIS方法，利用多模态预处理表示将HTE识别转化为马尔可夫毯发现问题，实现因果可解释的异质性处理效应识别，并在非洲反贫困项目中验证。

链接：https://arxiv.org/abs/2606.17010

机构：ISTA； UNICEF（联合国儿童基金会）； UCLA（加州大学洛杉矶分校）

作者：Riccardo Cadei, Frank Otchere, Nyasha Tirivayi, Gustavo Angeles Tagliaferro, Falco J. Bargagli-Stoffi, Francesco Locatello

英文摘要：Heterogeneous Treatment Effect (HTE) identification is crucial to explain the impact of an intervention and optimize our policies accordingly. Existing approaches trade expressivity for interpretability, but, if some active heterogeneity drivers are unmeasured, methods at both ends of this spectrum allow for spurious HTE characterization with no causal reading. In this work, we focus on controlled experiments and argue that an oracle HTE causal characterization via the latent interactors is now within reach, thanks to (i) more extensive pre-treatment measurements, i.e., multi-modal and multi-view, and (ii) scalable representations with minimal human supervision. We then re-frame HTE identification as a Markov-blanket discovery problem on a sufficient and aligned pre-treatment representation, and introduce Neural EXposure Interaction Search (NEXIS), an iterative procedure with provable and empirically validated consistent selection. We deploy NEXIS on two anti-poverty programs in Africa, augmenting each with satellite imagery capturing previously unmeasured environmental effect modifiers, leading to novel, interpretable and prescriptive guidelines to optimize the programs' next iterations.

AI 总结：本文通过LIBERO任务实验证明，按阶段局部应用度量进行演示筛选不如全局或统一度量，原因是缺陷信号被稀释且阶段度量不可迁移。

链接：https://arxiv.org/abs/2606.15064

机构：Department of Mechanical Engineering, University of California, Berkeley（加州大学伯克利分校机械工程系）

作者：Aarav Bedi

英文摘要：Manipulation demonstrations have temporal phase structure, and a natural hypothesis is that demonstration-curation metrics should be applied within phases rather than globally. The idea is to segment each trajectory into phases, score each phase with the metric that is locally most informative, and then aggregate. This follows directly from prior work showing that a single global metric can be the best detector of a defect and yet the worst curator of the resulting policy. We test the per-phase hypothesis on three contact-rich LIBERO pick-and-place tasks with a controlled early-release structural defect, comparing phase-gated curation against the same metrics applied uniformly and against a strong single global metric. Across all three tasks and five random seeds per condition, phase-gated curation is never the best curation strategy, and it is the worst of the three on two of the three tasks (Task 1: 86.0 vs. 92.0 for global; Task 3: 22.7 vs. 48.0 for uniform). We trace the failure to a concrete mechanism. When the defect signal is concentrated in a single phase, rank-aggregating across phases dilutes that signal with uninformative scores from defect-free phases, selecting a worse demonstration subset than simply applying the defect-informative metric everywhere. We further show that the per-phase metric selection does not transfer across tasks, since no phase shares a winning metric between any two tasks, so the selection cannot be reused and must be re-derived per task from a noisy sweep. These results bound a plausible and previously untested method, and they argue that practitioners should prefer identifying a single defect-informative metric over decomposing curation by phase. We release the full pipeline, all metric implementations, and per-seed results.

AI 总结：针对现有异常检测方法在流形支持数据上的局限性，提出基于投影算子的几何视角，将异常定义为投影残差，统一了重建方法并提升了性能。

链接：https://arxiv.org/abs/2606.15280

机构：Machine Learning Group, TU Berlin（柏林工业大学机器学习组）； BIFOLD, Berlin, Germany（柏林BIFOLD研究所）

作者：Alexander Bauer

英文摘要：Most existing anomaly detection methods rely on estimating a probability density or learning an enclosing decision boundary, implicitly assuming that normal data occupies a region of non-zero volume in the ambient space. In contrast, structural anomaly detection considers data that lies near a low-dimensional manifold, creating a mismatch between the inductive bias of existing methods and the structure of the data, often resulting in degraded performance. To address this mismatch, we introduce a geometric perspective. Specifically, we learn a projection operator onto the manifold of normal samples and define a sample as anomalous if it is altered by this projection. This formulation naturally integrates the inductive bias of manifold-supported data and reframes anomaly detection in terms of a projection residual, thereby resolving issues arising from modeling degenerate distributions. Notably, it provides a unifying interpretation of reconstruction-based methods by explaining their success and failure in terms of projection quality. In particular, it explains the strong generalization ability of projection-aligned models as a consequence of contraction behavior toward the manifold. Moreover, by decoupling anomaly detection from probabilistic modeling, it reduces the tendency to misclassify rare but normal samples, a widely recognized limitation of existing approaches. Empirically, we demonstrate that projection-aligned methods achieve strong performance, outperforming boundary-based methods while improving upon existing reconstruction-based approaches.

AI 总结：研究重复双边贸易中的公平性，提出Rawls-to-Nash公平增益目标族，并刻画其最优学习率。

链接：https://arxiv.org/abs/2606.15369

机构：University of Lille（里尔大学）； Institut Universitaire de France (IUF)（法国大学研究院）； School of Mathematics, University of Bristol（布里斯托大学数学学院）； Univ. Lille, CNRS, Inria, Centrale Lille, UMR 9189-CRIStAL（里尔大学、法国国家科学研究中心、法国国家信息与自动化研究所、中央理工-里尔高等电力学院，UMR 9189-CRIStAL）

作者：François Bachoc, Roberto Colomboni, Emilie Kaufmann

英文摘要：We study repeated bilateral trade from a fairness perspective. At each round, a fresh seller-buyer pair arrives, and the platform posts a price before observing the traders' valuations. Trade occurs only if both agents accept the price. Rather than maximizing only the gain from trade, we consider platforms that seek balanced divisions of the generated surplus. We show that natural fairness desiderata lead to a one-parameter Rawls-to-Nash family of fair-gain objectives, obtained by aggregating the seller's and buyer's net gains through nonpositive Hölder means. Unlike the standard gain-from-trade objective and the Rawlsian fair-gain objective studied in prior work, our proposed objectives induce a new statistical structure in which expected rewards are recovered from threshold feedback through a two-dimensional singular-kernel integral identity. This leads to a nonstandard pure-exploration problem whose natural estimators are rectangular double sums with row-column dependence and singular weights. Assuming independent i.i.d. seller and buyer valuation sequences with arbitrary unknown marginals, we characterize the optimal learning rates for the whole Rawls-to-Nash family of fair-gain objectives, giving matching fixed-confidence sample-complexity and regret bounds up to polylogarithmic factors.

AI 总结：提出开放智能的形式化定义，通过有限原始集和组合算子生成闭包，支持跨任务和世界的无限组合生成，并引入下一原始预测作为架构目标。

链接：https://arxiv.org/abs/2606.15386

作者：Ida Momennejad, Roberta Raileanu

英文摘要：Open-ended intelligence is the capacity to adapt to novel problems and environments that are substantially different from those in training. We formalize open-ended intelligence as the closure induced by a finite primitive set \(P\) and a set of composition operators \(C\). We characterize properties of the induced closure \(\mathcal{L}(P,C)\) that support unbounded compositional generation across families of tasks and worlds. A mathematics of open-ended intelligence requires two pillars: a minimal set of representational primitives (e.g., states, actions) and algorithmic primitives (e.g., nearest neighbor), together with composition motifs (e.g., recursion, sequencing) that reflect an acquired compositional grammar. The closure of these two pillars enables the generation of infinite adaptive responses across a wide range of settings. The mathematics supports complementary research agendas, including evaluation metrics for explanation and interpretability, as well as building architectures where compositional generalization is native. We propose next primitive prediction as a novel architectural objective, where the training objective encourages the acquisition of reusable algorithmic primitives and their compositional grammar, such that new solutions are generated through recombination. Curriculum learning and self-play enable lifelong learning and expansion of the closure by discovering reusable primitives and transition motifs across families of tasks and worlds. We ground the framework through case studies in physics, evolution, and neuroscience.

AI 总结：提出宪法价值潜力（CVP）方法，通过从隐藏状态学习标量势来读取模型内部的价值优先级边际，以预测和干预价值冲突，AUROC高达0.95。

链接：https://arxiv.org/abs/2606.15420

机构：NVIDIA Research（英伟达研究院）； Rutgers University（罗格斯大学）

作者：Tong Che, Rui Wu

英文摘要：A constitution tells a language model what to value, but little tells us whether it does. Adherence is judged from outputs, and output evidence is most fragile on value conflicts, where what matters is not which value a model mentions but which one it is willing to sacrifice. We provide evidence that this arbitration can be read from activations in a structured margin readout. We introduce Constitutional Value Potentials (CVP). For each value we learn a scalar potential from the hidden state: an internal pressure to preserve that value, supervised not by the prompt but by an independent judge's verdict on which value the model's own response actually preserved. The signed difference of two potentials is a priority margin. A constitutional clause becomes the claim that a margin stays positive, and a single monitor score flags when it does not. The monitor predicts conflict violations with AUROC up to 0.95, beats a strong hidden-state probe, and generalizes to held-out synthetic conflicts across three Qwen2.5 scales. The signal appears as the answer begins, from the prompt tail and first response token. Read this early, the same signal reveals whether an adversarial priority hack has actually pushed the model toward a violation, rather than only whether the prompt looks adversarial. The same directions also support intervention tests: under selected steering settings, moving along a value direction shifts judged trade-offs in the intended direction. Together, these results suggest that some constitution-relevant priorities are accessible as activation-space margins, rather than only as output behavior.

AI 总结：针对生成式AI中多阶段协作的价值分配问题，提出AME框架，整合异构数据贡献评估、数据权利映射和可信执行，实现与人类判断一致的低成本价值分配。

链接：https://arxiv.org/abs/2606.16075

机构：University of Shanghai for Science and Technology（上海理工大学）； Fudan University（复旦大学）

作者：Yang Shi, Songwen Pei, Yang Gao, Bingxue Zhang

英文摘要：Generative AI enables value creation through multi-stage collaboration among heterogeneous contributors, including training data, base models, fine-tuning behaviors, and prompts. However, how to fairly allocate the data value remains largely unexplored. This paper formulates multi-stage generative AI value allocation as a new research problem and identifies three core challenges: heterogeneous data contribution valuation, data rights mapping, and trustworthy execution. We propose AME (Attribution-Mapping-Execution) framework, a unified framework that integrates data contribution valuation, data rights mapping, and trustworthy execution into a single workflow. Experimental results demonstrate that AME framework achieves data value allocation outcomes more consistent with human reference judgments while maintaining low-cost trustworthy execution. Our work provides an initial foundation for value assessment and revenue allocation in generative AI data markets.

AI 总结：本文研究LLM电路发现中的重采样、重述和样本方差，提出CEAP方法减少重采样方差，并分析重述方差源于不同模板激活不同电路，样本方差主要由不忠定义导致。

链接：https://arxiv.org/abs/2606.16920

机构：Laboratory for Information and Inference Systems (LIONS), École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland（信息与推理系统实验室（LIONS），洛桑联邦理工学院（EPFL），瑞士洛桑）

作者：Frank Zhengqing Wu, Francesco Tonin, Volkan Cevher

英文摘要：Circuit discovery is a key technique in mechanistic interpretability to pinpoint the model components that are crucial for performing a given task. Although the current state-of-the-art method (EAP-IG) performs well on the metric of (un)faithfulness, it suffers from substantial variability. This includes resampling variance, where the circuit changes when we probe with a new batch of data from the same distribution; rephrasing variance, where the discovered circuit shifts when the prompts are rephrased; and sample-wise variance, where a circuit with low population unfaithfulness exhibits large fluctuations in unfaithfulness across individual samples. This paper studies the roots of these variances. We demonstrate that CEAP, our new circuit discovery method that improves upon EAP-IG with a theoretical guarantee, can substantially lessen resampling variance. We further show that rephrasing variance arises because prompts with different templates tend to activate different circuits in the model. This leads us to argue that it may be challenging to find a comprehensive circuit that explains and controls the model's behavior on a task, which can be expressed in countless templates, suggesting that LLMs may be inherently hard to steer. We show that sparsity, which has been claimed to form more compact and interpretable task circuits, fails to solve this problem. Regarding sample-wise variance, we argue that it is largely benign: extremely poor unfaithfulness scores often stem from how unfaithfulness is defined, rather than from defects in the measured circuits. We show that the magnitude of unfaithfulness is affected by selective contribution scaling, a neural mechanism that accounts for the extremely poor scores sometimes observed.