机器学习学术速递[7.31]

AI 总结：本文针对小数据工程代理建模的过拟合与资源开销问题，评估三种递归Transformer架构，验证其在两项任务中可实现预测精度、参数效率与计算成本的有效权衡。

链接：https://arxiv.org/abs/2607.27251

作者：Kart-leong Lim

英文摘要：Transformer-based surrogate models are increasingly used to replace expensive first-principles simulation in engineering design. But conventional transformer architectures are often over parameterized for the small, low-dimensional datasets typical of engineering design spaces, where large simulation data is expensive to generate. Under these conditions, excess parameter capacity leads to overfitting rather than improved accuracy, while also incurring unnecessary memory and compute overhead. This motivates a shift towards architectures that focus on additional compute rather than additional learnable parameters. This paper presents a hardware-aware evaluation of three recursive transformer paradigms for surrogate thermo-mechanical analysis of advanced packages: a)Tiny Recursive Model, b) our proposed Depth Recursive transformer, c) and a simple recursive transformer. We systematically compare their predictive performance (Recall, Mean Reciprocal Rank), parameter count, computational complexity (FLOPs), providing practical design guidelines for selecting recursive transformer architectures under resource-constrained scenarios. We validate this principle on two low-dimensional engineering prediction tasks: 1) thermo-mechanical reliability analysis of advanced semiconductor packages, where stress and warpage from thermal cycling must be evaluated repeatedly across a design-of-experiments sweep under costly finite element analysis (FEA). 2) Laplace PDE iterative numerical solver for capacitance field. Overall, recursive weight-sharing transformers provide an effective and generalizable trade-off between prediction accuracy, parameter efficiency, and computational cost for small data engineering surrogate modeling.

AI 总结：针对大语言模型后训练中数据组织固定导致的样本优化失衡问题，提出SDO框架，通过逐轮调整小批量与样本暴露，加快多任务收敛并均衡不同问题类型的准确率。

链接：https://arxiv.org/abs/2607.27273

作者：Jinliang Gao, Ning Yang, Hai Wang, Baili Xiao, Pin Lyu

英文摘要：Post-training of large language models is expensive, and existing efficiency improvements mainly focus on selecting informative samples or designing training schedules. However, data organization itself is usually treated as a static preprocessing step: embedding-based grouping methods construct fixed partitions before training and cannot adapt to the evolving sample exposure during optimization. As a result, all samples receive similar exposure despite their different optimization needs, leading to redundant updates for some samples while leaving others under-optimized. To address this problem, we propose SDO (Structure-Aware Data Organization), a plug-and-play data organization framework with an exposure-driven feedback mechanism that organizes mini-batch composition and sample exposure according to representation-space structure. SDO operates epoch by epoch on frozen external embeddings, avoiding model warm-up training overhead: within each epoch, locality-aware batching forms coherent mini-batches via KNN neighborhood traversal; across epochs, exposure-balanced scheduling records per-sample participation and reduces the sampling probability of over-exposed samples to preserve long-term coverage. Across SFT, DPO, and GRPO, SDO accelerates convergence, with the largest gains observed in the early-to-mid phase, producing more coherent gradients and more balanced accuracy across question types without permanently excluding training samples.

AI 总结：本文以τ²-bench为基准，发现量化大语言模型智能体的分数看似平稳，但会放大原有工具调用失败，其损害被基准的10个错误预算掩盖，缩小预算可暴露该问题，相关诊断方法可与任务奖励一同报告。

链接：https://arxiv.org/abs/2607.27275

作者：Jiwon Jang, Kisu Yang, Heuiseok Lim, Hyunwoo Park

英文摘要： Post-training quantization to 4-bit weights is widely reported to be nearly lossless. We test this claim for multi-turn, tool-calling agents, where it now matters most. On $\tau^2$-bench, across two open-weight model families in dense and MoE variants and two domains (eight cells, 456 episodes each, at 16-, 8-, and 4-bit weights), quantization indeed looks free on the standard metric. No cell shows a score change that survives multiple-comparison correction, and in the cell that carries the largest process damage, equivalence testing bounds the change within $\pm$7.5 points. The process tells a different story. Quantization amplifies the failure the model already exhibits at full precision (tool-name hallucination in telecom, with the same directional trend in retail entity errors) by up to 2.5$\times$ in volume (+17.6 points per task), while creating essentially no new failures. The failure set is the same at every precision (rank correlation $\geq$ 0.94, 0.18% novel events). The score stays flat because the benchmark's ten-error budget absorbs the extra failures. Shrinking the budget to two errors re-exposes a score gap of 17 points, and it does so only in the one cell where quantization added error volume, exactly as the masking account predicts. A targeted error-repair prompt, run for five telecom models at every precision, removes the damage exactly and only where it lives. Both diagnostics, the per-channel error rate and success under a shrinking budget, come from logs benchmarks already collect; we suggest reporting them alongside task reward.

AI 总结：该研究提出TIER-MoE风险引导子空间混合专家模型，用于生物医学分类多模态融合，可提升预测性能与概率校准，在多数据集上优于现有最优方法且具备强零样本泛化能力。

链接：https://arxiv.org/abs/2607.27289

作者：Yu Chang, Anzhe Cheng, Chenwei Wu, Zhuoran Wang, Jiahao Chen, Tamoghna Chattopadhyay, Sophia I. Thomopoulos, Paul M. Thompson, Liyue Shen, Paul Bogdan

英文摘要：The promise of multimodal fusion lies in combining complementary sources of evidence, yet more evidence does not always yield a better prediction. Recent multimodal models have advanced fusion through richer cross-modal interaction and sample-adaptive fusion. However, the influence assigned to a modality during fusion does not reveal whether that source is unreliable, redundant, or poorly matched to a specialized expert. To address this limitation, we introduce TIER-MoE, a risk-guided subspace mixture-of-experts model that defines sample-specific modality reliability as the prediction loss its unimodal predictor is expected to incur. This risk is learned from out-of-fold predictions generated by models that were not trained on the corresponding sample. TIER-MoE combines the estimated risk with expert-specific subspace compatibility for sparse modality-expert routing, while an always-active shared path preserves multimodal complementarity. We evaluate TIER-MoE on four public multimodal biomedical datasets spanning Alzheimer's disease status, skin-lesion malignancy, and retinal classification. Results demonstrate its superiority over state-of-the-art methods in predictive performance and probability calibration, with consistent improvements in Macro-F1 and Brier score and strong zero-shot generalization to an external cohort.

AI 总结：该研究提出Conditional Informer架构，通过条件注意力机制编码船舶与环境的物理依赖，结合Modality Masking策略，在AIS与ERA5数据上提升了船舶轨迹预测准确率并降低了传感器故障时的误差。

链接：https://arxiv.org/abs/2607.27418

机构：Carnegie Mellon University（卡内基梅隆大学）； Advanced Technology Laboratories Lockheed Martin（洛克希德·马丁公司先进技术实验室）

作者：Yuan Guan, Chandler Squires, Timothy Hu, Pradeep Ravikumar

英文摘要：Long-term ship trajectory prediction is a fundamental capability for maritime safety and autonomous navigation. While recent Transformer-based architectures have improved forecasting horizons, they predominantly rely on historical kinematic states, treating vessel motion as an isolated system. In reality, maritime navigation is profoundly modulated by extrinsic factors like weather and constrained by static vessel characteristics. Existing multimodal approaches fundamentally model the joint distribution over states and contexts, treating environmental variables as peer features rather than encoding the directional physical dependence of vessel dynamics on environmental conditions. In this work, we propose the Conditional Informer, a novel encoder-decoder architecture that formulates trajectory prediction as a conditional generation task. We employ a dedicated Conditional Attention mechanism where the vessel state explicitly queries environmental contexts through cross-attention, encoding the physical prior that weather modulates - but is not generated by - vessel dynamics. Furthermore, to address the intermittency of real-world data, we introduce a Modality Masking training strategy to prevent catastrophic degradation during sensor fallback. Extensive experiments on AIS and ERA5 data demonstrate that our approach outperforms kinematic and concatenation-based baselines by 15.4% in prediction accuracy when context is available. Crucially, Modality Masking prevents shortcut learning, reducing fallback error by nearly an order of magnitude compared to unconstrained models.

AI 总结：该研究针对语言模型记忆的精确删除问题，将删除方式分为减法和重放，在不同参数规模的Gemma 3和Kimi Linear混合模型中验证了两种方式的效果，明确了精确删除与记忆表示的关联。

链接：https://arxiv.org/abs/2607.27539

机构：Vy Labs, Inc.（Vy Labs公司）

作者：Vishwajith Ramesh

英文摘要：Exact deletion from persistent language-model memory depends on how that memory represents a record. Addressable influence can be removed by algebraic decrement; influence transformed by later writes inside shared recurrent state requires rebuilding from before the write. We test this distinction in two pretrained models against explicit record-omitted references. First, we replace Gemma 3's global-attention layers with support-vector memory. After low-rank recovery at 1B, decrement and retained-key refit agree at the next-token output to median KL $5.4\times10^{-15}$ over 31 support-token deletions, with $+2.0\%$ perplexity relative to a matched fine-tune. A masked-refit proxy is indistinguishable from the never-ingested floor under elicitation, relearning, sampling, and LiRA attacks. At 4B and 12B, certificate ordering persists but utility cost rises to $11.2\%$ and $44.3\%$. Second, in a 48B Kimi Linear hybrid, additive writes admit a fixed decrement and diagonal decay a corrected one, whereas the delta rule makes $12$--$49\%$ of a record's contribution suffix-dependent. Checkpointed rewind-and-replay deletes real clinical records at contexts up to 18,842 tokens, matching never-ingested logits and all recurrent states bit for bit within a deterministic MLX implementation; replaying a correction provides exact amendment. Exact deletion is therefore a property of memory representation: subtract addressable records and replay entangled writes.

AI 总结：本文提出KGPS方法，将提示选择转化为动态状态估计问题，适配RL训练动态，在多推理基准和RL算法上较基线提升准确率与rollout效率，在在线提示选择中达最优。

链接：https://arxiv.org/abs/2607.27610

机构：Beihang University（北京航空航天大学）； Zhongguancun Academy（中关村学院）； Nanyang Technological University（南洋理工大学）； Communication University of China（中国传媒大学）

作者：Haodong Zhu, Yangyang Ren, Yanjing Li, Sheng Xu, Haiguang Liu, Linlin Yang, Baochang Zhang

英文摘要：Reinforcement learning (RL) finetuning significantly enhances the reasoning capabilities of large language models (LLMs), yet its effectiveness critically depends on selecting prompts of appropriate difficulty for the current policy. This is challenging because prompt difficulty evolves throughout training. Existing online methods therefore face a trade-off: evaluation-based approaches are accurate but expensive, while prediction-based approaches are efficient but typically assume stationary difficulty, making them ill-suited to RL's non-stationary training dynamics. To address these issues, we propose a Kalman-Guided Prompt Selection method (KGPS), which reformulates prompt selection as a dynamic state estimation problem rather than static difficulty prediction. KGPS models each prompt's latent success rate in logit space using a linear-Gaussian state-space model, with process noise coupled to the magnitude of policy updates so that uncertainty increases when the policy changes more substantially. A Kalman filter then maintains a calibrated Gaussian posterior over prompt difficulty, and prompts are selected by maximizing a posterior-expected training utility that favors intermediate-difficulty prompts while naturally revisiting uncertain ones. The resulting procedure is adaptive to policy drift and requires no additional rollouts beyond standard policy training. Extensive experiments across mathematics, planning, and geometry reasoning benchmarks, as well as multiple RL algorithms, show that KGPS consistently improves both final accuracy and rollout efficiency over strong baselines, establishing state-of-the-art performance among online prompt selection methods. For example, on DeepSeek-R1-Distill-7B, KGPS uses 83% fewer rollouts than DS while even improving the average performance by 0.12 point across six math reasoning benchmarks.

AI 总结：该研究从凸优化视角推导了适用于通用优化器和模型架构的损失联合表征，得到闭式最优批量大小调度与联合缩放律，其性能优于固定批量大小基线，凸显动态批量大小调度在大语言模型训练中的重要性。

链接：https://arxiv.org/abs/2607.27731

机构：Meta

作者：Jiaxiang Li, Zhiqi Bu, Shiyun Xu

英文摘要：Modern deep learning typically keeps the batch size static throughout training, thus overlooking the joint effect of learning rate and batch size on the training dynamics. In this paper, we study the deep learning dynamics through the lens of convex optimization and derive a joint characterization of loss in terms of both schedules, applicable to general optimizers and model architectures. This characterization yields a closed-form optimal batch size schedule for any prescribed learning rate schedule, and further leads to joint scaling laws that consistently outperform static batch size baselines, highlighting the significance of dynamic batch size schedule in large language model training.

AI 总结：该研究针对连续EEG监测中Transformer架构的内存瓶颈，提出S-CEReBrO架构，通过窗口交替注意力实现内存恒定，在EEG下游任务中性能领先且效率提升。

链接：https://arxiv.org/abs/2607.27913

机构：ETH Zurich（苏黎世联邦理工学院）； Nanyang Technological University（南洋理工大学）； University of Bologna（博洛尼亚大学）

作者：Glenn Anta Bucagu, Thorir Mar Ingolfsson, Yawei Li, Luca Benini

英文摘要：Foundation models offer a promising paradigm for Electroencephalography (EEG) analysis, leveraging generalizable representations from vast unlabeled datasets. Yet, Transformer-based architectures face a critical bottleneck: global attention mechanisms couple the attention memory state to the signal duration, causing memory overflow during continuous monitoring. To address this, we introduce S-CEReBrO (Streaming CEReBrO), an evolution of the CEReBrO architecture designed for continuous monitoring. Our novel Windowed Alternating Attention mechanism factorizes attention computation into fixed-size spatiotemporal windows, guaranteeing constant KV cache memory as only the active window requires resident attention maps. Empirical scaling analysis confirms that windowed alternating attention can process signals 100X longer than full self-attention and 3X longer than low-rank linear attention. Compared to low-rank linear attention on long contexts, windowed alternating attention requires 55% of the memory while increasing inference throughput by 2.1X. Pre-trained on >25,000 hours of recordings from >12,000 subjects, S-CEReBrO achieves state-of-the-art performance on 7 of 11 downstream tasks, with up to 60% fewer parameters. This work represents a significant step toward the realization of efficient, generalizable, and continuous EEG monitoring. An accompanying code repository is available.

AI 总结：该研究发现稀疏MoE中专家约简顺序等因素会导致执行行为分歧，确定了post-mHC与完整持久状态的边界属性，为MoE运行时和硬件后端提供了数值兼容性契约。

链接：https://arxiv.org/abs/2607.28097

作者：Tianyang Zhu

英文摘要： Mathematically equivalent expert-reduction orders can produce observably different sparse-MoE executions. We isolate this effect in native DeepSeek-V4-Flash by freezing local MoE state and varying only aggregation semantics. Four schemes separate operand representation from accumulator precision. At one layer-5 fork, 720 A-mode orders yield 10 continuation basins; 720 B-mode orders form 360 exact structural classes and 11 basins. Under one Chinese prompt, the B classes split into 202 layoffs, 113 hiring, and 45 other continuations. Maximum-L-infinity B-branch selection separates 12, 24, and 36 of 50 prompts by 8, 16, and 32 tokens. Across 192 persistent trajectories per scheme, P32, A, and B change every native-reference route trajectory, while C preserves routes, token sequences, and texts. A separate 192-trajectory C check matches native MoE, post-mHC, next-router, and LM states bitwise. For one controlled B branch, exact post-mHC endpoint reconstruction reproduces the measured downstream trajectory. At the next decode boundary, exact FP64 reconstruction of the branch's full persistent state yields agreement for 301 downstream post-mHC states, 301 persistent-state checkpoints, 301 routes, predictions, and text over seven steps, given the same naturally generated next input. These controls identify post-mHC as an intra-token boundary and full persistent state as a cross-token continuation boundary. Identical tokens need not imply identical autoregressive state: divergence can survive a token boundary and become visible later. These results make expert operand conversion, accumulator precision, and reduction order part of a numerical compatibility contract for sparse-MoE runtimes and hardware backends. They establish controlled causal possibility, not deployment incidence; C's order invariance is limited to evaluated six-term states and schedules.

AI 总结：该研究区分了MoE路由的相干性等量，提出相干重叠概念，发现所选专家子空间重叠显著但实际路由更优，且添加后续专家多能提升预测，说明几何相似性无法决定冗余或剪枝价值。

链接：https://arxiv.org/abs/2607.28308

作者：Huiyuan Tian, Bonan Xu, Shijian Li

英文摘要：Sparse mixture-of-experts (MoE) language models route each token to multiple experts, suggesting a geometric account of their benefit: co-selected experts should contribute distinct representation directions. Existing evidence often conflates route coherence, candidate quality, and candidate-by-context interaction. We distinguish these quantities using an Expert Subspace Separation Index (ESSI), matched-route residuals, and a prefix-controlled $2\times2$ factorial; frozen-route interventions and a controlled Top-$k$ study assess functional value. Three paired contrasts organize the findings. First, across six MoE architectures, expert subspaces overlap substantially, yet actual routes explain token representations better than matched alternatives. Second, across the 39 factorial cells in OLMoE, Mixtral, and DeepSeek, the selected candidate explains more of the residual representation than the strongest unselected rival in every cell, yet the actual prefix narrows this advantage throughout: all interactions are negative, and every 95% confidence interval lies below zero. Third, this geometric narrowing does not imply functional redundancy: adding later experts improves next-token prediction in 24 of 39 frozen-route comparisons, while the other 15 estimates are inconclusive; a controlled training study also favors Top-2 over Top-1 in all three seeds. We call this joint pattern coherent overlap: routing selects token-relevant experts from a shared geometric neighborhood, while useful multi-expert computation persists without disjoint linear coverage. Separating these quantities clarifies why geometric similarity alone cannot determine redundancy or pruning value.

AI 总结：针对EEG疾病诊断中实例标签可靠性不均、受试者稀缺的问题，提出两阶段BridgeMIL框架，解耦实例表示学习与受试者级监督，在15种数据集-骨干网络设置中14种获最高准确率，总体较基线高4.28个百分点。

链接：https://arxiv.org/abs/2607.27274

作者：Zhiyuan Ma, Zeyuan Li, Zhiyi Lu, Jiacheng Hao, Youlang Du, Zhen Jiang, Xinche Zhang, Yuhao Sun, Sen Song

英文摘要：EEG-based disease diagnosis requires one prediction per subject, yet common pipelines segment recordings into short instances, inherit the subject label for every instance, and train instance-level classifiers. This assumes that all instances provide equally reliable diagnostic evidence. Multiple instance learning (MIL) avoids inherited labels by treating each subject as a bag. However, EEG datasets contain far fewer subjects than instances, which can limit the quality of the representations learned by end-to-end MIL. We propose BridgeMIL, a two-stage framework that decouples instance representation learning from subject-level supervision. Stage 1 pretrains the encoder without inherited instance labels by aligning temporally nearby windows and independently sampled within-subject sub-bags. Variance and covariance regularization prevent collapse and reduce redundancy without negative pairs. Stage 2 transfers the encoder to an attention-based MIL aggregator, applies supervision only to subject predictions, and limits representation drift through feature retention. Across three EEG disease datasets and five representative backbones, BridgeMIL attains the highest mean accuracy in 14 of 15 dataset-backbone settings and an overall mean accuracy of 76.57%, 4.28 percentage points higher than the strongest baseline. Further analyses reveal substantial variation in inherited-label reliability across instances, greater performance sensitivity to subject scarcity than to instance scarcity, and a more structured representation space with distinct subject-wise clusters and improved separation between diagnostic classes. Together, these findings underscore the importance of aligning supervision with the subject-level prediction objective while learning from abundant EEG instances without assigning disease labels to individual instances.

AI 总结：本文针对一般实值矩阵三分解，开展稀疏诱导可识别性的严格理论研究，提出新颖分解策略推导恢复保证等结果，经蒙特卡洛实验验证，填补了相关理论空白。

链接：https://arxiv.org/abs/2607.27507

作者：Tingting Mu

英文摘要：Matrix factorisation is a fundamental tool for exploiting low-dimensional structure in high-dimensional data, with applications such as data compression, denoising, structure discovery, interpretable representation learning, and dimensionality reduction. Compared to conventional two-factor models, matrix tri-factorisation provides greater modelling flexibility, while sparsity constraints often improve both interpretability and recovery performance. Although the role of sparsity has been extensively studied for two-factor matrix factorisation, rigorous theoretical guarantees for general real-valued matrix tri-factorisation remain largely unexplored. To address this gap, we establish, to the best of our knowledge, the first rigorous theoretical study for sparsity-induced identifiability in general real-valued matrix tri-factorisation. Our analysis is enabled by a novel decomposition strategy that transforms the original problem into two coupled auxiliary factorisation problems, while preserving the structural information necessary to the recovery of the original factor matrices from the observations. Building upon this decomposition, we derive recovery guarantees and structural consistency results that characterise how coefficient sparsity influences the sufficient recovery conditions, convergence behaviour, spectral approximation error, high-probability bounds, and structure preservation. Comprehensive Monte Carlo experiments validate the proposed theory and demonstrate close agreement between the theoretical results and empirical observations.

AI 总结：本研究构建统一理论框架揭示子模信息测度（SIMs）的几何与统计特性，通过控制合成实验验证，为选择设计基于SIMs的表示学习目标提供原则性指导。

链接：https://arxiv.org/abs/2607.27660

机构：The University of Texas at Dallas（德克萨斯大学达拉斯分校）； Adobe（奥多比公司）

作者：Rishabh Iyer, Truong Pham, Anay Majee

英文摘要：Submodular Information Measures (SIMs) have recently emerged as a powerful framework for representation learning and multimodal learning. In particular, the SCORE framework~\cite{majee2024score} demonstrated that SIMs can serve as effective objectives for supervised contrastive learning. Despite their empirical success, however, the geometric and statistical properties induced by different submodular information measures remain poorly understood. In this work, we develop a unified theoretical framework connecting SIMs to classical concepts in representation learning and statistical pattern recognition. We show that Total Information (TI) objectives characterize intra-class structure: Graph Cut TI recovers within-class variance, LogDet TI recovers generalized variance and covariance volume, and Facility Location TI induces imbalance-aware separation that emphasizes rare and confusable classes. We further show that Mutual Information (MI) objectives capture complementary notions of inter-class structure: Graph Cut MI is closely related to centroid separation and Fisher-style discrimination, LogDet MI captures covariance-aware separation through Mahalanobis distance, and Facility Location MI measures nearest-mode representational overlap. We validate these theoretical characterizations using controlled synthetic experiments that independently vary variance, covariance, class imbalance, class separation, and multimodal overlap. Across all settings, the empirical behavior closely matches the proposed theory. Our results provide the first unified geometric and statistical understanding of submodular information measures and offer principled guidance for selecting and designing SIM-based objectives for representation learning.

AI 总结：该研究针对多视图聚类中的部分视图对齐问题，提出DAS-PMVC框架，通过锚点图结构对齐、结构增强特征学习与双对齐策略提升性能，在多数据集上优于现有最先进方法。

链接：https://arxiv.org/abs/2607.27761

机构：Dalian University of Technology（大连理工大学）； Inspur Group Co., Ltd.（浪潮集团有限公司）； China University of Mining and Technology（中国矿业大学）； The University of Warwick（华威大学）

作者：Shubin Ma, Liang Zhao, Chuanye He, Zhenjiao Liu, Liang Zou, Lin Yuanbo Wu, Yu Shao

英文摘要：In recent years, multi-view clustering has attracted widespread research interest. However, due to limitations in data collection devices, data across different views often suffer from misalignment, leading to the partial view alignment problem (PVAP). To mitigate the impact of view asymmetry and irrelevant samples, this paper proposes a framework for partial multi-view clustering via dual alignment and structure enhancement (DAS-PMVC), which leverages view structure consistency and semantic relevance. Specifically, DAS-PMVC includes three parts: \textbf{anchor graph structure alignment}, where sample joint embedding representations with consistent latent space are derived from anchor point relationships for initial view alignment; \textbf{structure-enhanced feature learning}, where the model learns view structure information through pretraining and combines multi-view graph convolutional networks to further extract deep latent features from the aligned graph structure to improve the discriminative power of representations; and \textbf{a dual alignment strategy}, where initial alignment is performed through the anchor graph in the pretraining phase, and contrastive learning loss and the Hungarian algorithm are introduced in the training phase to further optimize the alignment of latent features. Experimental results on various datasets demonstrate that the DAS-PMVC framework outperforms existing state-of-the-art methods in clustering performance, showcasing its effectiveness and superiority.

AI 总结：该研究针对开放网络RTB场景下用户身份碎片化问题，构建基于自监督学习的用户基础模型，经LLM优化预训练后，在竞价胜率、CTR等生产任务中取得显著性能提升。

链接：https://arxiv.org/abs/2607.28019

机构：Teads（泰兹）

作者：Solal Vernier, Ivan Can Arisoy, Merwan Barlier, Blaž Škrlj

英文摘要：User foundation models have demonstrated strong results in e-commerce and social recommendation, but most industrial deployments assume environments where user identity is stable and persistent. Open-web real-time bidding (RTB) operates on a structurally different data distribution: user identity is fragmented and non-persistent across browsing sessions, and the availability of browsing history depends on user privacy choices. Consequently, a significant portion of traffic carries no historical data, and available records often consist of relatively short, disjointed sessions. As a result, historical signals in this domain are typically represented as aggregated counters and recency buckets, leaving the sequential structure unexploited. To address this limitation, we present a user foundation model that applies self-supervised learning on user browsing histories and show that the learned representation improves multiple downstream production tasks, demonstrating the viability of this approach on the open web. We pre-train a Transformer encoder with masked language modeling and a sequence-level contrastive objective, then fine-tune it on the click prediction task. We optimize the encoder's pre-training pipeline with an LLM-in-the-loop search over a curated catalog of reviewable, code-level edits (lifters), instantiating the LLM-as-optimizer paradigm in an industrial setting. The same encoder representation yields +1.197% RIG on the production bid win-rate model and +1.354% RIG on the production CTR ranker; a 7-day live A/B test confirms +2.13% CTR, -1.13% eCPC (80% CI excluding zero on both metrics).

AI 总结：该研究针对在线策略自蒸馏的暴露偏差问题，提出CRPO方法，通过对比学习与分组对比保留优化信号，在13个推理基准上提升了训练稳定性与泛化性。

链接：https://arxiv.org/abs/2607.28026

作者：Xingjian Wu, Junlin Liu, Xingchen Liu, Xuhang Zhu, Jianing Wang, Linsen Guo, Xiaoyu Li, Xuezhi Cao, Xunliang Cai

英文摘要： Recent advances in post-training Large Language Models (LLMs) increasingly rely on Reinforcement Learning with Verifiable Rewards (RLVR) or On-Policy Self-Distillation (OPSD). While OPSD provides dense, logit-level supervision, it inherently suffers from exposure bias due to the privileged information of the self-teacher. In multi-turn agentic settings, this leads to reasoning route convergence and the loss of clear optimization directions. To tackle these challenges, we introduce Contrastive Reinforced Policy Optimization (CRPO), which reformulates agentic OPSD from a contrastive learning perspective. By leveraging predictive entropy to distinguish between positive positions (reflective exploration) and negative positions (exposure bias), CRPO conducts group-wise contrast to preserve reliable, fine-grained optimization signals. Extensive evaluations across 13 challenging reasoning and deep-search benchmarks demonstrate that CRPO consistently outperforms existing reinforcement learning and self-distillation baselines, significantly enhancing training stability and generalization in long-horizon interactions.

AI 总结：该研究提出RLPF方法，将执行结果转化为分阶段奖励，微调Qwen3-32B后提升了代码生成的正确性与效率，模型性能可适度迁移，验证了代码智能体可优化生成程序。

链接：https://arxiv.org/abs/2607.27271

机构：HKUST（香港科技大学）

作者：Huihao Jing, Haozhe Cui, Wenbin Hu, Shaojin Chen, Haochen Shi, Changxuan Fan, Yuxuan Liu, Hanyu Yang, Sirui Zhang, Ziyi Chen, Haoran Li, Yangqiu Song

英文摘要：Code models are increasingly trained with execution feedback, but most training signals still stop at correctness. This leaves an important gap for systems code: two programs can pass the same tests while differing greatly in runtime. We study how to train code agents to prefer faster correct implementations, rather than treating efficiency only as an evaluation metric. The key difficulty is that runtime is a fragile reward. It is meaningful only after a program is correct, varies across tasks, and gives little guidance when most sampled programs fail to compile or run. We propose \textbf{RLPF}, reinforcement learning from performance feedback, which turns execution outcomes into a staged reward. Failed programs are ordered by execution progress, while correct programs are ranked by their relative improvement from the baseline toward the expert reference. This gives useful feedback before correctness and performance-sensitive feedback after correctness. Fine-tuning Qwen3-32B with RLPF on PerfCodeBench raises correct-and-runnable solutions from $11.1\%$ to $54.6\%$ and improves relative efficiency from $8.1\%$ to $38.6\%$. The trained model becomes competitive with stronger open-weight systems, and its optimization behavior transfers modestly to EffiBench-X. Additional studies show that model-generated references provide useful but weaker supervision, and that the full composite reward is more reliable than correctness-only or runtime-only baselines. These results suggest that code agents can be trained not only to pass tests, but also to optimize the programs they write.

AI 总结：针对现有激活引导无法引导双路径网格世界简单策略的问题，提出PGS方法，经网格世界、国际象棋、竞技足球实验验证其校准性、可逆性及行为适配的可组合性与跨域迁移性。

链接：https://arxiv.org/abs/2607.27574

作者：Yoann Poupart, Aurélie Beynier, Nicolas Maudet

英文摘要：Activation steering has emerged in large language models as a lightweight alternative for dynamically changing a model's behavior at inference time. However, we show that existing steering methods fail to steer even a simple policy in a two-route gridworld environment. To address this limitation, we propose Policy Gradient Steering (PGS), which formulates steering as a reinforcement learning problem. PGS accumulates gradients of a temporary behavioral objective over a small set of rollouts or demonstrations to construct a removable task vector. We first demonstrate the calibration and reversibility of PGS in a two-route gridworld environment. Using chess puzzles, we then evaluate independently fitted PGS vectors both in isolation and in combination, finding that compatible tactical objectives accumulate constructively. Finally, in competitive football, we show that PGS can alter specific team behaviors and that its effects transfer across opponents. Together, these results show that policy gradients provide a natural interface for constructing temporary and composable behavioral adaptations across diverse decision-making domains.

AI 总结：该研究提出“先扩展后压缩”框架，结合RGRPO与TU-OPD等技术，构建互补教师并集并压缩，使Qwen3-1.7B学生模型在数学推理等三领域优于最强单教师，实现性能提升。

链接：https://arxiv.org/abs/2607.27770

机构：Moore Threads AI（摩尔线程人工智能）

作者：Songshuo Lu, Zhi Chen, Yaohua Tang

英文摘要：A single reinforcement-learning run can produce a strong reasoner yet an incomplete teacher: it often amplifies only a subset of the valid solution modes. We argue that reinforcement learning (RL)-trained policies should therefore be viewed as local probes of a multi-basin reasoning solution manifold, rather than as globally reliable supervisors. Based on this view, we propose an expand-then-compress framework that couples teacher construction with multi-teacher policy distillation. In the expansion stage, Residual Group Relative Policy Optimization (RGRPO) trains a sequence of teachers from a common initialization and redirects each later round toward examples not yet covered by the accumulated teacher union. In the compression stage, reliability-gated Teacher-Union On-policy Distillation (TU-OPD) lets the student learn from its own response prefixes. For each example, only reliable teachers contribute, and their sampled-token OPD losses are weighted by their per-example quality. We further introduce Consensus-Residual Decomposition, which preserves a winner teacher's excess token preferences over its reliable peers, preventing specialist behavior from being suppressed during teacher aggregation. Experiments on mathematical reasoning, code generation, and instruction following show that the resulting Qwen3-1.7B student consistently outperforms the strongest individual teacher across all three domains, yielding relative improvements of 2.0%, 8.3%, and 6.9%, respectively, while retaining single-model inference. These results establish a simple but powerful principle: stronger students can be obtained not by selecting a single better teacher, but by deliberately constructing and compressing a complementary teacher union.

AI 总结：针对数学推理 RLVR 的悬崖提示梯度丢失问题，提出 LSPO 采样时机制，通过 LoRA 适配器恢复梯度，在 DeepMath-103K 数据集上 16 项基准指标均优于 DAPO，平均提升 3.8 分。

链接：https://arxiv.org/abs/2607.27787

机构：NVIDIA（英伟达）

作者：Ken Ding

英文摘要：Reinforcement learning from verifiable rewards (RLVR) for mathematical reasoning suffers from a structural blind spot: on "cliff" prompts-those on which every sampled rollout in a group fails-the group-normalized advantage is identically zero, so GRPO produces no gradient on precisely the prompts at the frontier of the model's capability. We introduce LoRA Scaffolded Policy Optimization (LSPO), a sampling-time mechanism that recovers this lost gradient. Each RL step, LSPO detects cliff prompts, fits a small low-rank (LoRA) adapter by a brief supervised step on their ground-truth solutions, re-rolls the cliffs with the base-plus-adapter model, splices the now-successful completions back into the RL batch with an importance-sampling correction, and takes a GRPO step on the base alone; the adapter receives only the supervised gradient and is discarded at checkpoint, yielding a base-only model. On DeepMath-103K with DeepSeek-R1-Distill-Qwen-1.5B, evaluated over n=5 paired seeds per arm at a matched 1000-step reporting horizon, LSPO's 5-seed mean matches or beats a DAPO baseline on all 16 (benchmark, pass@k) cells (15 strict wins and one exact tie), with gains of up to +10.7 points on AIME24/pass@4, +6.7 points on AIME24 and AIME26 at pass@16, and +2.4 points on MATH500/pass@1; averaged over the 16 cells the improvement is +3.8 points.

AI 总结：本研究提出类别感知强化学习方法，将实例预测类别纳入RL状态表示，经7个不同领域数据集对比实验，该方法收敛更快、奖励优化更好、有效反事实解释更多，凸显类别感知的重要性。

链接：https://arxiv.org/abs/2607.27905

机构：Institute of Business Administration Karachi（卡拉奇工商管理学院）； Commonwealth Bank of Australia（澳大利亚联邦银行）； Stanford University（斯坦福大学）； University of New South Wales（新南威尔士大学）

作者：Muhammad Adil Saleem, Syed Ali Raza, Mary-Anne Williams

英文摘要：Counterfactual explanations (CFEs) enhance the interpretability of black-box models by generating alternative instances with adjusted feature values that achieve a contrastive outcome. Reinforcement learning (RL) offers a promising approach for CFE generation, enabling efficient exploration of counterfactual instances while ensuring control over key metrics like validity, sparsity, and proximity. Previous studies have formulated RL states exclusively using features derived from the predictors in the supervised dataset. This study explores the impact of including an instance's predicted class, alongside features derived from the predictors, in the RL state representation for generating CFEs. The hypothesis is that class-awareness enhances exploration efficiency and improves policy optimality. We compare the proposed class-aware RL method with the class-blind RL method, which is similar but excludes the instance's class information from the state representation. The comparison was conducted using seven datasets from diverse domains, varying in size. The results show that during training, class-aware RL offers benefits in terms of convergence speed, reward optimization, and episode length reduction. Moreover, it generates significantly more valid CFEs compared to class-blind RL. Finally, the instance's class-based feature consistently ranks among the most influential predictors in RL's action-selection, as shown by the SHAP and LIME values, underscoring the significance of class-awareness in RL for CFE generation. The impact is heightened clarity, faster learning, improved validity, and more effective counterfactual generation across diverse datasets.

AI 总结：该研究针对多区域VAV控制，测试前沿LLM的控制能力及TD3引导的RFT效果，发现RFT未产生持续改进，需先开展状态转移聚焦的监督微调。

链接：https://arxiv.org/abs/2607.27914

机构：The University of Tokyo（东京大学）； Tokyo University of Science（东京理科大学）

作者：Takumi Shioda, Kohei Terashima, Tatsuo Nagai

英文摘要：Multi-zone variable-air-volume control must balance thermal comfort, indoor air quality, and electricity use across several continuous actuators. Model predictive control and reinforcement learning are widely studied, but deployment typically requires building-specific modeling or training, limiting scalability. We first test whether a frontier reasoning model (an LLM trained to use additional inference-time computation) can achieve competitive VAV control from text without building-specific training. With that capability established, we then test whether TD3-guided reinforcement fine-tuning (RFT) can transfer control knowledge into a locally deployable open-weight model. Five controllers are evaluated over three summer days in a physics-based four-zone emulator. Relative to a Guideline 36-based baseline, TD3 reduced HVAC electricity by 4.5% while improving temperature and CO$_2$ compliance. Without building-specific training, GPT-5 achieved the largest reduction (6.2%) but reduced the ventilation margin. For RFT, deterministic rollouts restore a saved state, apply one candidate, and follow TD3 to score each action. Auditing a learned critic against these rollouts exposed a failure hidden by its near-perfect across-time correlation ($r=0.9998$): within-state ranking was unreliable; the critic selected the rollout-best candidate in only 5 of 10 states. Even with the rollout verifier, 200 RFT steps produced no sustained improvement in sampled-action return; the open-weight controller used more electricity than the baseline before and after training, and its five-minute predictions remained worse than persistence. GPT-5 predicted transitions far better. Exact rollout scores rank sampled actions but reveal neither next-state effects nor an improvement direction. The unchanged transition errors motivate transition-focused supervised fine-tuning before value-based RFT.

AI 总结：该研究提出首个LLM引导的AutoPref框架，将偏好目标分解为成对损失与集合感知加权程序，通过分阶段条件搜索策略实现自动发现，在TSP等四类组合优化问题上性能优于手动设计基线。

链接：https://arxiv.org/abs/2607.27953

作者：Shengda Gu, Kai Li, Xinyi Ke, Haobo Fu, Yifan Zhang, Jian Cheng

英文摘要：Combinatorial optimization problems (COPs) underpin many real-world decisions, but their exponentially large search spaces make high-quality solutions costly to obtain. Neural combinatorial optimization (NCO) learns fast construction policies, typically with reinforcement learning (RL), while preference-based NCO improves sample efficiency by learning from relative solution quality. However, existing preference objectives combine two distinct design choices in manually specified, one-size-fits-all formulations: what learning signal to extract from each solution pair and how to weight each pair relative to the sampled set. We present AutoPref, the first LLM-guided framework for automated preference-objective discovery in NCO. AutoPref factorizes the objective into a pairwise loss program, which defines the learning signal, and a set-aware weighting program, which determines each pair's relative contribution. Their composition forms a unified programmatic objective space containing existing preference objectives as special cases. To make its search tractable, we introduce a staged conditional search strategy with behavioral gates that filter inadmissible programs before short-horizon training and evaluation. Across TSP, CVRP, FFSP, and JSSP, AutoPref consistently outperforms strong hand-designed baselines across problem scales, demonstrating the benefits and scalability of automated objective discovery for NCO.

AI 总结：该研究在RUL框架下提出两种新型奖励函数，经RWKU基准实验，其遗忘效率较二元奖励提升，速度快达3倍且保留模型效用，表明奖励设计是机器遗忘效率的关键驱动因素。

链接：https://arxiv.org/abs/2607.27968

机构：Technical University of Munich（慕尼黑工业大学）； Sapienza University of Rome（罗马大学）； Munich Center for Machine Learning（慕尼黑机器学习中心）

作者：Efstratios Zaradoukas, Davide Gabrielli, Bardh Prenkaj, Gjergji Kasneci

英文摘要：Machine unlearning seeks to selectively remove specific knowledge from trained language models without full retraining, a growing necessity under privacy regulations such as GDPR and the EU AI Act. Recent work has reformulated unlearning as a Reinforcement Learning with Verifiable Rewards (RLVR) problem, where models are optimized against verifiable rewards computed directly from their outputs. However, existing methods rely on sparse binary rewards that provide minimal learning signal, indicating only whether forbidden content was avoided, and limiting convergence speed. In this paper, we study how reward design affects unlearning efficiency within the Reinforcement Unlearning (RUL) framework. We introduce a principled reward decomposition framework that decouples verifiability from sparsity, and propose two new reward functions: an exponential reward that provides graded penalties based on the count of forbidden-concept occurrences, and a PageRank inspired reward that weights penalties by semantic importance. We conduct experiments on the Real World Knowledge Unlearning (RWKU) benchmark, demonstrating that both rewards consistently outperform the binary setting, while reaching similar forgetting performance up to $3\times$ faster and preserving general model utility. Our results show that reward design is a key driver of unlearning efficiency offering a practical path toward scalable and efficient machine unlearning.

AI 总结：该研究提出TAPO框架，通过策略优化与过渡监督交替训练，增强LLM智能体对环境过渡的敏感性，作为轻量即插即用模块，在WebShop和ALFWorld实验中提升任务性能。

链接：https://arxiv.org/abs/2607.27973

机构：School of Electronic and Computer Engineering, Peking University（北京大学电子与计算机工程学院）； Pengcheng Laboratory（鹏城实验室）

作者：Cong Li, Peixi Peng, Yisen Zhao, Xinyu Hu, Shudong Liu, Zhan Su, Zhuojian Li

英文摘要：Recently, Reinforcement Learning (RL) has emerged as a crucial paradigm for the post-training of Large Language Model (LLM) agents. However, existing methods predominantly rely on sparse task rewards for policy optimization, failing to fully exploit another class of inherently dense supervisory signals naturally present during online interaction: environmental feedback following action execution. Recent theoretical studies suggest that generalization in multi-step, goal-oriented tasks hinges on predictive knowledge of environmental consequences. Inspired by this, we propose TAPO: Transition-Aware Policy Optimization for LLM Agents, a unified training framework that alternates between policy optimization and transition supervision. Beyond standard RL updates, TAPO repurposes rollout data to apply action-conditioned next-observation prediction supervision on a shared backbone model. This approach enhances the model's sensitivity to environmental transition dynamics and action consequences while concurrently optimizing the policy. It serves as a computationally lightweight, plug-and-play enhancement module for existing agent RL algorithms, requiring no additional expert data, extra sampling costs, or inference-time overhead. We conduct systematic experiments on WebShop and ALFWorld, integrating foundation models of various scales with different policy optimization algorithms. Empirical results demonstrate that TAPO consistently improves task performance over pure policy optimization baselines.

AI 总结：针对复杂文档与会议场景文本转SVG的三大挑战，提出GVR-Coder框架，引入定制数据集DocMeetSVG-100K，结合课程微调、双渲染反馈强化学习与智能体循环，生成质量优于基线。

链接：https://arxiv.org/abs/2607.28073

机构：University of Science and Technology of China（中国科学技术大学）； ByteDance Inc.（字节跳动公司）； Hefei University of Technology（合肥工业大学）

作者：Yiming Xu, Jihua Kang, Chunsai Du, Qifan Zhang, Wangqiu Zhou, Yiting Wu, Tianqi Li, Qi Song

英文摘要： In demanding professional environments and meeting review scenarios, lengthy text often imposes a high cognitive load. To facilitate efficient information communication, transforming verbose text into logically clear diagrams is essential. Scalable Vector Graphics (SVG) provide an effective representation for this purpose due to their editability and resolution independence. However, current research on Text-to-SVG generation remains hindered by three major challenges: (1) the scarcity of datasets for complex, logic-rich diagrams; (2) the absence of explicit layout priors, which leads to chaotic spatial arrangements; and (3) the lack of fine-grained visual feedback to validate rendered outputs and correct aesthetic defects. To address these challenges, at the data level, we introduce DocMeetSVG-100K, a large-scale SVG dataset tailored for document authoring and meeting review scenarios. At the model level, we propose GVR-Coder, a novel framework designed to generate high-quality logical diagrams from lengthy professional texts. Specifically, we adopt a curriculum-driven rejection sampling fine-tuning to progressively enhance the model's capability in modeling complex structures, while explicitly incorporating layout constraint knowledge during training. In addition, we introduce reinforcement learning from dual rendering feedback, a mechanism that provides implicit feedback through reward signals to jointly optimize structural complexity and visual aesthetics. Furthermore, we design a generate-verify-repair agent loop, which improves generation quality through explicit, fine-grained feedback and targeted refinement. Extensive experiments demonstrate that GVR-Coder outperforms competitive baselines and reliably produces logically coherent and visually appealing diagrams. Code and data are available at this https URL.

AI 总结：本研究提出LM-GRASP框架，将GRASP随机构造阶段转为在线模仿学习任务，用仅解码器Transformer作构造策略，在Taillard PFSP基准上平均比GPU-GRASP提升28.4个制造跨度单位，是手工构造器的实用替代方案。

链接：https://arxiv.org/abs/2607.28135

作者：Mohand Mezmaz, Grégoire Danoy

英文摘要：Machine learning for combinatorial optimization typically relies on neural constructors trained via reinforcement learning on large offline datasets for a fixed problem class-incurring high pretraining costs and generalizing poorly outside the training distribution. We propose an alternative: a metaheuristic framework that reformulates the randomized constructive phase of GRASP as an online imitation learning task, trained from scratch on each problem instance. A local search procedure acts as an expert oracle, while a decoder-only Transformer serves as the constructive policy. Unlike classical GRASP, which relies on static, myopic heuristic rules based on localized scalar costs, our approach is fully data-driven: the construction policy emerges from high-quality solutions discovered during the search itself, with no problem-specific feature engineering required. We instantiate this as LM-GRASP, a hybrid metaheuristic following an iterative learn-infer-improve cycle, training the policy online via behavioral cloning on a dynamic archive of elite trajectories-no external data or offline pretraining needed. The pipeline interfaces with the domain solely through the objective evaluator used by local search. Evaluated on the Taillard PFSP benchmark (ta51-ta60), the most discriminating block due to half its optima being unknown, LM-GRASP outperforms GPU-GRASP by 28.4 makespan units on average-comparable to the gain from GPU acceleration over sequential execution (27.2 units), though with overlapping standard deviations. This suggests instance-specific, online-trained language models are a promising, practical alternative to hand-engineered constructors, especially for landscapes resistant to classical greedy construction.

AI 总结：针对HPC任务的LLM RL后训练的异质性问题，提出HARGO方法，通过置信度调制优势实现逐响应重要性加权，在四项HPC任务的三项核心指标上均取得最优性能。

链接：https://arxiv.org/abs/2607.28301

作者：Tiangang Li, Xiangbo Tian

英文摘要：Supervised fine-tuning (SFT) can equip large language models (LLMs) with domain knowledge for high-performance computing (HPC) tasks such as data race detection and benchmark question answering. However, knowledge alone does not guarantee task-appropriate behavior: the same SFT model that correctly classifies 88.65\% of C/C++ data race samples produces verbose, imprecise answers to factual queries, with 65.9\% of MLPerf responses exceeding 40 characters. Reinforcement learning (RL) post-training addresses this gap by optimizing for task-specific rewards rather than token-level imitation. Yet HPC tasks exhibit extreme heterogeneity, with binary classification, factual QA, and semantic generation differing by 58x in answer length, spanning three distinct reward distributions, and showing widely varying SFT accuracy. This makes uniform-weight RL methods such as GRPO suboptimal. We propose HARGO, Heterogeneity-Aware Reward-Guided Optimization, which introduces per-response importance weighting via confidence-modulated advantage: computing a discrimination signal from group-level reward contrast and a confidence signal from reference model log-probabilities, then modulating the advantage before computing per-response weights, without requiring task-type labels. Across four HPC tasks and nine methods, HARGO achieves the best performance on all three primary metrics: WinRate 54.62\%, Data Race F1 91.30\%, and PLP Similarity 0.8558. Ablation confirms complementary contributions from both signals. HARGO establishes the best overall alignment quality among compared methods for heterogeneous HPC tasks.

AI 总结：该研究针对平均回报CMDP中神经评论家的阶最优收敛问题，提出分层MLMC神经评论家，开发出实现阶最优的原始-对偶自然演员-评论家算法，无需知晓混合时间。

链接：https://arxiv.org/abs/2607.28390

作者：Ankur Naskar, Vaneet Aggarwal

英文摘要：Constrained Markov Decision Processes (CMDPs) provide a natural framework for reinforcement learning in safety-critical applications, where agents maximize long-term reward while satisfying long-term constraints. Although primal-dual actor-critic methods with linear critics are well understood, extending order-optimal convergence guarantees to neural critics in average-reward CMDPs has remained open. The main challenge is a fundamental bias-cost trade-off in neural critic estimation: under Neural Tangent Kernel (NTK) analysis, reducing critic bias substantially increases critic optimization cost, preventing order-optimal convergence in the primal-dual framework. We resolve this bottleneck by introducing a hierarchical Multilevel Monte Carlo (MLMC) neural critic that performs debiasing simultaneously across trajectory sampling and critic optimization. The resulting estimator attains the bias of a long critic optimization run with only logarithmic expected sample cost. Building on this estimator, we develop a primal-dual Natural Actor-Critic algorithm that achieves both an optimality gap and a constraint violation of order $\tilde{O}(T^{-1/2})$. This establishes the first order-optimal convergence guarantees for infinite-horizon average-reward CMDPs with general policy parameterization and neural critics, while eliminating the need to know the underlying mixing time. Our results are novel even in the unconstrained setting.

AI 总结：本论文针对大动作空间交互式系统的策略学习挑战，提出meTS、dTS、sDM等结构化贝叶斯方法，解决在线与离线策略学习的关键问题并提供理论保证。

链接：https://arxiv.org/abs/2607.28408

作者：Imad Aouali

英文摘要：This thesis studies policy learning in interactive systems where an agent observes a context, selects an action from a very large set, and receives partial feedback. The main framework is contextual bandits, with two paradigms: on-policy learning, where the agent interacts sequentially with the environment and minimizes regret, and off-policy learning, where it learns from logged data collected by a logging policy. In large action spaces, both settings face major challenges: inefficient exploration, sparse data coverage, high-variance importance weights, extrapolation bias, and difficult optimization landscapes. The first part develops structured Bayesian methods for on-policy learning. We introduce meTS, a mixed-effect extension of Thompson sampling, and dTS, which leverages diffusion-inspired priors to model dependencies between actions. These methods share information across actions and yield regret guarantees depending on an effective number of actions. The second part addresses off-policy learning. We propose sDM, a structured direct method based on latent variables, show that optimization error can dominate estimation error in large action spaces, and introduce concave, efficiently optimizable policy-weighted log-likelihood objectives. Finally, we develop differentiable pessimistic methods based on exponential smoothing and PAC-Bayesian bounds to control the bias-variance trade-off of regularized importance-sampling estimators.

AI 总结：该研究针对SOC警报疲劳问题，训练了思维链推理增强的分类器及校准器，在Windows终端检测任务上提升了良性与恶意召回率，且30B微调模型优于通用大模型。

链接：https://arxiv.org/abs/2607.28460

作者：Amol Khanna, Manu Nandan, Cristian Viorel Popa, Joan Pujol-Roig, Diana Bolocan, Laura Vasilie, Alexandru Apostu, Chase Helwig, Mihaela Gaman, Michael Brautbar, Edward Raff, Chase Midler, Sven Krasser

英文摘要：A major issue in Security Operations Centers (SOCs) is alert fatigue, as the number of detections reported is more than staff can triage in a given day. Prior work prompts or fine-tunes large language models (LLMs) to emit a triage label directly, but does not train them to reason about whether a detection is a genuine threat. We train a chain-of-thought (CoT) reasoning-enabled triage classifier on real, human-labeled Windows endpoint detections by combining automated prompt optimization, self-training, and reinforcement learning with verifiable rewards. We find that CoT reasoning also degrades the label-token probabilities that automated triage relies on, so we separately train a calibrator that reads the full reasoning trace and estimates the probability that the verdict is correct. Our system reaches 82.6% test accuracy and, at the high-confidence operating point that governs automated triage, improves benign recall by 43.0% and malicious recall by 18.3% over a direct-label LLM classifier. We further show that the trained calibrator is necessary - an untrained confidence judge collapses high-confidence recall to zero - and that a finetuned 30B model significantly outperforms frontier general-purpose models, motivating targeted training over scale.

AI 总结：β-OPSD是将普通OPSD扩展为含可控正则化参数β的策略优化通用形式，通过蒸馏近似策略优化解，在数学推理基准上性能与稳定性均优于普通OPSD。

链接：https://arxiv.org/abs/2607.28582

机构：University of Maryland（马里兰大学）

作者：Jiawei Xu, Minghui Liu, Juzheng Zhang, Tom Goldstein, Furong Huang

英文摘要：On-policy self-distillation (OPSD) is a promising approach to improve reasoning language models, but it remains brittle in practice: making it work reliably often requires substantial engineering effort. We identify a structural source of this difficulty: vanilla OPSD is precisely the $\beta=1$ member of a broader policy-optimization family, where $\beta$ weights the KL penalty anchoring the student to a reference policy. This equivalence turns $\beta$ from an implicit value fixed at one into a controllable regularization parameter, yielding a more general formulation that trades off proximity to a reference policy against privileged teacher guidance. We introduce $\beta$-OPSD and derive its optimal policy as a geometric interpolation between the reference policy and the privileged teacher. Directly optimizing this objective with reinforcement learning, however, would be costly and high-variance. Rather than optimize the RL objective directly, we turn its closed-form solution into a distillation target. Each value of $\beta$ selects a target along the reference-to-teacher path, which we implement efficiently by mixing their token-level logits. In this way, inexpensive distillation approximates the solution of expensive policy optimization. Return-to-go credit assignment further aligns token updates with the sequence-level objective while retaining the simplicity of OPSD. Experiments on mathematical reasoning benchmarks show that $\beta$-OPSD consistently outperforms vanilla OPSD, improving optimization stability and downstream reasoning performance. Our results provide a principled route from self-distillation to policy optimization and back without sacrificing the efficiency that makes OPSD practical.

AI 总结：DoTime是用于干预和反事实时间序列的合成基准生成器，具备现有工具没有的多项功能，附带评估套件与基线，经测试其干预训练的因果模型相比同容量观测模型有方向准确性优势。

链接：https://arxiv.org/abs/2607.27263

机构：National University of Singapore（新加坡国立大学）

作者：Dennis Thumm, Billy Tim Anthony, Ying Chen

英文摘要： Most benchmarks for causal inference over time series are observational, small, or domain-specific, leaving interventional and counterfactual estimation under-served exactly where it matters most, such as in healthcare, policy evaluation, and climate science. We introduce \textbf{DoTime}, an open, scalable, and theoretically grounded generator of multivariate temporal structural causal models (TSCMs) with interventions, released as the \code{dotime} PyPI package together with four frozen evaluation suites. Beyond existing work, it adds capabilities absent from prior generators: continuous-time intervention \emph{windows}, counterfactual sampling modes with a positivity guard, regime-switching SCMs as a strict generalization of interrupted time series, non-stationary dynamics by construction with switching SCM parameters, and deterministic ramp and sinusoidal intervention profiles that place trends and structural breaks \emph{inside} the evaluation window. Moreover, it demonstrates the suitability of the generator as a prior for a causal foundation model reference implementation. The released suites span a training-scale snapshot of $100{,}000$ trajectories and eight named identification structures, each with exact ground truth: paired interventional trajectories from the same SCM throughout, and shared-noise counterfactuals in the continuous-time suite. We ship reference baseline implementations with an evaluation harness, and pose a falsifiable claim: interventional training buys a measurable direction-accuracy advantage over an observational model of identical capacity. It is tested across three training seeds per arm. Under structure-matched evaluation on held-out episodes, the interventional prior-fitted network's (PFN) gap is positive in every structure, trajectory length, and seed tested.

AI 总结：该研究针对现有自动竞价算法仅聚焦DSP侧的问题，提出首个以统一广告平台为中心的自动竞价基准PlatformBid，定义三类竞争场景并评估多种方法，还提出BidFlow方法，在快手实验中目标成本提升0.68%。

链接：https://arxiv.org/abs/2607.27265

机构：Southeast University（东南大学）； Kuaishou Technology（快手科技）； Nanyang Technological University（南洋理工大学）

作者：Shengtian Yang, Yewen Li, Peng Jiang, Zhiyi Lyu, Bo An, Peng Jiang, Qingpeng Cai, Lei Feng

英文摘要：Real-time bidding is central to computational advertising, comprising three elements: Supply Side Platform (SSP) selling ad impressions, Demand Side Platform (DSP) bidding for advertisers, and Ad Exchange conducting auctions between them. Traditional auto-bidding algorithms focus solely on the DSP side, maximizing advertiser conversions by adjusting bids against competitors. However, current big ad platforms, such as social media and e-commerce companies, now integrate SSP, DSP, and Ad Exchange functions internally. From such ad platforms' perspective, the goal of the auto-bidding algorithms is not only to maximize the advertisers' conversions, but also the total revenue of the platform. Given the lack of platform-centric evaluation frameworks and the pressing need to advance auto-bidding research, we propose PlatformBid - the first comprehensive benchmark designed from a unified ad platform's perspective. To accurately reflect the real-world auto-bidding scenarios, we define three representative settings: (1) homogeneous competition with identical algorithms across advertisers, (2) heterogeneous competition with diverse algorithmic strategies, and (3) promotional competition where some advertisers surge budgets for boosting sales during promotional events like Black Friday. We systematically evaluate a broad spectrum of existing auto-bidding methods across these settings, encompassing classical control methods, RL-based methods, and recent generative methods. Besides these methods, we further propose a novel auto-bidding method based on flow-matching, termed BidFlow, which leverages the flow-matching method's expressive policy representation to effectively handle dynamic competitive environments. Online experiments on Kuaishou further show a +0.68\% improvement in target cost, providing deployment evidence for the offline-online consistency of PlatformBid.

AI 总结：EvoCause 用 LLM 优化因果图以提升电信等系统的根本原因分析性能，发布了 TeleRCA 基准，在合成数据和真实网络数据上均优于基线方法。

链接：https://arxiv.org/abs/2607.27290

作者：Lei Zan, Keli Zhang, Shifeng Xie, Jiale Zheng, Zehao Xiao, Zhiwei Dong, Ke Zhang, Ruichu Cai, Malik Tiomoko, Lujia Pan

英文摘要：Modern telecommunication, cloud, and microservice systems emit correlated alarm cascades when components fail. Root cause analysis (RCA) aims to identify the small set of alarms that initiate each cascade. A common approach learns a causal graph from observational logs and predicts all zero-in-degree alarms in each incident-induced subgraph. However, the learned graph remains fixed and cannot benefit from expert diagnoses of historical incidents. We close this loop with EvoCause. Expert labels constrain which alarms should be source nodes but do not specify the edge edits needed to satisfy those constraints. EvoCause uses a large language model (LLM) to propose semantically plausible graph edits, while deterministic code validates node identities and acyclicity and retains the best graph on a labeled alignment set. At test time, the refined graph alone produces transparent predictions without an LLM call. We also release TeleRCA, an expert-annotated benchmark from a production telecommunication network containing $485{,}681$ alarm events spanning $194$ alarm types over $5{,}621$ resources. On synthetic data, EvoCause initialized with the PC causal discovery algorithm outperforms the unrefined PC baseline, raising Node F1, Case EM, and Graph F1 by $11.59$, $9.40$, and $4.59$ percentage points, respectively, while reducing nSHD by $0.2379$. On TeleRCA, replacing human-readable alarm titles with anonymous identifiers lowers Node F1 and Case EM by $6.12$ and $8.04$ percentage points, respectively, indicating that alarm-name information contributes to graph refinement.

AI 总结：SE(3)-MeanFlow将MeanFlow扩展到李群几何，通过新训练目标实现少步蛋白质主链生成，性能优于多倍采样的流匹配基线，适配高通量设计需求。

链接：https://arxiv.org/abs/2607.27431

机构：Purdue University（普渡大学）； Vanderbilt University（范德堡大学）； Freie Universität Berlin（柏林自由大学）

作者：Yikun Bai, Binghang Lu, Yikai Liu, Elaheh Akbari, Soheil Kolouri, Linxuan Wang, Ping He, Shuchan Wang, Ruqi Zhang, Guang Lin

英文摘要： Generative modeling of protein backbones promises the de novo design of proteins with prescribed structural and functional properties. Existing diffusion and flow-matching models produce high-quality backbones on SE(3)^N, but inference requires numerically integrating an ODE over hundreds of network evaluations, each involving a Lie group exponential map - a bottleneck for high-throughput design campaigns. We introduce SE(3)-MeanFlow, a few-step generative framework that extends MeanFlow from Euclidean space to the Lie group geometry of protein frames. Working natively in the Lie algebra so(3) and in R^3, we derive closed-form average-velocity identities for rotations and translations, giving simulation-free training targets. We further introduce an SE(3) alpha-Flow objective that removes the Jacobian-vector product from the rotation branch and serves as a warm-up stage, after which training switches to a small-t stabilized MeanFlow loss that is used for the remainder of pretraining and for rectification-based post-training. In protein backbone generation, SE(3)-MeanFlow matches or exceeds flow-matching baselines that use several times more sampling steps, and its advantage widens in the few-step regime, where rectification lets it lead at every matched budget - at a modest cost in diversity.

AI 总结：提出隐核离散流图（LKF）模型，通过共享隐变量绑定分解组件实现少步生成，在两个文本基准上较似然基线提升困惑度2.1-3.3倍，M=8时优于现有少步采样器

链接：https://arxiv.org/abs/2607.27529

作者：Mansoor Ahmed, Yue-Tsz Fan, Hemanth Venkateswara, Murray Patterson

英文摘要：Discrete diffusion and flow-matching models denoise a sequence over many steps, but to keep each step cheap, they factorize the transition across positions and decide every token independently. This makes few-step generation challenging for text when the target couples two positions, such as a subject and a verb that must agree. An independent update commits to them separately, and many function evaluations are spent repairing the mismatch. Existing few-step methods buy back the lost correlation by distilling or rectifying a slow teacher, and so inherit the teacher's quality ceiling. We ask instead whether a model can express correlated steps natively, and answer with Latent-Kernel Discrete Flow Maps (LKF), a from-scratch flow-map kernel that is a mixture of M factorized components tied by a single shared latent. Conditioned on the latent, each component is cheap, and the mixture is summed over the latent in closed form for small M. We show that a single step places mass on correlated completions with the same sampling time complexity as a factorized model, since one latent is drawn per sequence and reused across the entire denoising trajectory. We also show that the Masked Diffusion Language Model (MDLM) is a special case of our LKF model at M=1. The experiments for unconditional text generation on the One-Billion-Word (LM1B) and WikiText-103 benchmarks show that our LKF model learns strongly heterogeneous components and improves generative perplexity by 2.1x to 3.3x over the likelihood baselines without losing diversity. The gain grows with M, and at M=8, it surpasses distilled and rectified few-step samplers. The source code is available at: this https URL

AI 总结：该研究提出受物理学启发的KSSE模型，将预训练特征映射到稀疏图并优化拓扑，在ImageNet-1000直推式任务中以更少参数实现高准确率，优于Swin-L、匹配ViT-H/14并减少模型占用。

链接：https://arxiv.org/abs/2607.28428

机构：South-West State University (SWSU)（西南州立大学（SWSU））； T8 LLC（T8有限责任公司）

作者：V.S. Usatyuk, D. A. Sapozhnikov, S. I. Egorov

英文摘要：We introduce Kohn--Sham Spectral Embedding (KSSE), a physics-inspired energy-based model replacing dense CNN classifiers with a sparse-graph spectral embedding evaluated at the Nishimori temperature of an associated Random-Bond Ising Model. By mapping pre-trained features onto quasi-cyclic low-density parity-check graphs and constructing a regularized Laplacian acting as a Kohn--Sham Hamiltonian, we solve $D$ independent channel spectral problems in $\mathcal{O}(N\log N + k^2_{\text{mode}} N)$ time via FFT on circulant blocks (leveraging Pontryagin self-duality of $\mathbb{Z}/p\mathbb{Z}$) and low-order Rayleigh refinement. Graph topology is optimized using \emph{star-domain surgery}: rather than destroying information-carrying codewords by removing frustrated cycles, we construct edge shifts creating local convexity around codewords while bounding residual frustration to $\rho(B_\gamma)\leq 1+\delta$. Multi-scale fractal analysis ($D_2$ spectrum) and fractal learning-rate landscape certifies a landscape transition from rough regimes ($D_2>3$) to star-domain basins ($D_2<1$), enabling Rayleigh refinement with $k_{\text{mode}}=5$ modes. We prove six theoretical results: a generalized Ihara--Bass identity linking belief propagation to the Laplacian; trapping-set eigenvalue correspondence; additive channel separability with an explicit exchange-correlation bound; a surgery theorem bounding frustration with attractor width $\Omega(1/\sqrt{d_{\min}})$; a quasi-stationarity perturbation bound; and a fixed-point convergence theorem. In a transductive protocol on ImageNet-1000 with frozen EfficientNet-B4 features ($D=1792$), KSSE achieves \textbf{88.93\%} Top-1 accuracy using $\approx 21.24$M parameters, outperforming Swin-L (197M, 86.4--87.3\%) and matching ViT-H/14 (632M, 88.0--89.5\%) under standard inductive setups, while reducing model footprint by $10\times$ and $30\times$, respectively.

AI 总结：本研究针对算法追索仅关注翻转模型预测的缺陷，提出因果表演框架建模追索行动的因果传播，实验显示其优于标准方法并减少模型重训需求。

链接：https://arxiv.org/abs/2607.28497

机构：Georgia Institute of Technology（佐治亚理工学院）； Vector Institute（向量研究院）； Drexel University（德雷塞尔大学）； Leiden University（莱顿大学）

作者：Srikanth Avasarala, Varun Gupta, Shahin Jabbari, Saber Salehkaleybar, Juba Ziani

英文摘要：Algorithmic recourse aims to provide individuals with actionable changes to improve their predicted outcomes in high-stakes classification settings, such as loan and mortgage applications. However, most existing approaches focus only on flipping a model's prediction, without accounting for whether the recommended changes lead to genuine improvement in an individual's true qualifications or merely enable strategic gaming of the classifier. Consequently, deployed recourse policies can induce behavioral responses that degrade predictive accuracy and become ineffective after model retraining. In this work, we formalize this failure mode through a causal performative framework for recourse. We model how recourse actions propagate through a structural causal model, capturing interactions among features as well as their effect on the true label. These causal responses induce a non-convex optimization problem, even under standard convex losses. We characterize conditions under which performatively stable solutions exist and can be efficiently computed via simple iterative dynamics. Our analysis reveals that recourse policies that ignore causal structure can induce large, misaligned behavioral responses, whereas causal recourse leads to stable equilibria that reduce incentives for gaming. Experiments on both semi-synthetic and real credit datasets demonstrate that our approach consistently outperforms standard empirical risk minimization while reducing the need for repeated model retraining to accommodate distribution shifts caused by strategic agent behavior.

AI 总结：本研究提出无监督原子策略优化框架APO，通过双奖励机制实现原子系统三维结构预测，在晶体与抗体预测中优于全监督基线，提升了匹配率、结构保真度与推理效率。

链接：https://arxiv.org/abs/2607.28553

机构：CMU（卡内基梅隆大学）； NUS（新加坡国立大学）

作者：Shentong Mo, Yatao Bian

英文摘要：Predicting the 3D structures of atomic systems is fundamental to advancing material science and drug discovery. While flow-matching models (, FlowDPO) have recently shown promise in this domain, their performance relies heavily on alignment with ground-truth coordinates via supervised preference learning. However, obtaining experimental labels for novel crystal phases or de novo proteins is prohibitively expensive, creating a bottleneck for structural modeling in data-scarce regimes. In this work, we propose (Atomic Policy Optimization), a fully unsupervised alignment framework that eliminates the need for ground-truth reference structures. APO adapts group-relative policy optimization to 3D atomic environments, utilizing a novel dual-reward mechanism: (i) a that reinforces the policy's dominant latent structural modes through eigen-decomposition of sample similarities, and (ii) a that enforces thermodynamic stability. Our framework enables the model to ``self-correct'' by identifying physically plausible configurations within sampled groups. Extensive benchmarks on crystal and antibody structure prediction demonstrate that APO consistently outperforms fully supervised baselines, achieving a new state-of-the-art in match rates and structural fidelity. Furthermore, we show that APO effectively straightens probability paths, significantly improving inference efficiency. Our results suggest that intrinsic physical consistency can serve as a superior guide for alignment compared to noisy, supervised coordinate matching.

AI 总结：该研究提出语言模型能力形成的成核速率定律，揭示无部分 credit 的联合对齐是速率限制步骤，定位可塑性丧失的损伤并找到重新初始化 query-key 切片的解决方案，适用于14亿参数 transformer 的合取回路。

链接：https://arxiv.org/abs/2607.27281

作者：Lei Dong

英文摘要：A capability appears in a language model when the last parts of its circuit align in one stochastic attempt, and getting all but one right is worth nothing. We show this no-partial-credit joint alignment is the rate-limiting step of capability formation. Two fingerprints: in a shortcut-free apparatus a five-part circuit missing three waits as long as a three-part circuit missing three (1.19-1.37), so the wait counts missing parts, not size; and on Pythia across seven capabilities and three scales, ablating one part leaves a median 17% of the capability in 32 of 32 discriminating cells, where partial credit predicts 50-83% (p = 2e-10), while a random non-part head leaves 100%. One rare event whose barrier grows with missing parts yields a rate equation -- sites x attempts x drive x exp(-beta*K), minus destruction -- read three ways, each preregistered with frozen constants. Forward: a capability flat at baseline ignites at a step of our choosing once the mix passes a concentration floor (10/10 above, 0/12 below), and while still flat its arrival is datable from its precursor to 5% median error on six held-out models. Backward: the delay to learn a withheld capability grows with waiting until, past a critical step, it never ignites -- yet validation loss falls smoothly throughout, so standard monitors are blind to it. We locate the damage (heads commit to the base data) and isolate the cure: re-initializing only the query-key slices restores learnability (6/6) while the value slices do nothing (0/6). We prove the mechanism in a controlled gated-attention model: occupation forces a deadline whose consequences need no mixing assumption. Completed: SGD's noise fails the fluctuation-dissipation test, so we install one and anneal, melt and pin circuits on schedule. Scope: conjunction circuits in transformers to 1.4B.

AI 总结：研究重尾噪声下Adam优化器的收敛性，推广在线到非凸转换框架分析Adam，明确其在重尾噪声下的收敛特性、次优性及域半径控制下的收敛速率提升情况。

链接：https://arxiv.org/abs/2607.27383

机构：Michigan State University（密歇根州立大学）

作者：Yijiang Pang

英文摘要：We establish the first convergence guarantees for the plain vector-form \emph{Adam} optimizer under heavy-tailed stochastic noise. While several Adam variants are known to achieve optimal iteration complexity in bounded-variance nonconvex optimization, little is understood about their behavior when stochastic gradients admit only a bounded $p$-th central moment for some $p \in (1,2]$, a setting increasingly observed in modern deep learning. To address this gap, we generalize the recent online-to-nonconvex conversion framework to accommodate heavy-tailed martingale-difference noise. Building on this generalized framework, we develop a discounted regret analysis for Adam, without restrictive parameter coupling. Our results show that Adam converges to $(\rho,\epsilon)$-stationary points under heavy-tailed noise. However, it exhibits a suboptimal iteration complexity and $p$-dependent convergence, a suboptimality that persists even in the bounded-variance case ($p=2$). When the domain radius is known and used to control the online-learner output, a standard setup in related literature, the convergence rate improves to match the optimal complexity. These findings provide new theoretical insight into the robustness and limitations of Adam in heavy-tailed regimes.

AI 总结：针对安全关键物联网系统的硬峰值信息年龄截止期限问题，提出OCO-PAoI-Hard算法，通过将调度转化为约束在线凸优化，实现零建模状态违规，性能优于多个基准方法。

链接：https://arxiv.org/abs/2607.27626

作者：Wentao Zhang, Wentao Mo

英文摘要： Safety-critical IoT systems such as industrial closed-loop control, V2X coordination, and remote teleoperation require every sensor's peak Age of Information (peak AoI, also abbreviated PAoI) to stay below a hard per-slot deadline, not merely an average bound. Existing approaches meet this requirement only under restrictive assumptions: stochastic channels for Whittle-index AoI, simulator rollouts for deep reinforcement learning, or sublinear cumulative violation for long-term constrained online convex optimization. Under adversarial coefficients, OCO-PAoI-Hard guarantees zero per-slot violation of the modeled AoI state under one-step viability and O(sqrt(T)) regret against any static safe comparator; packet-level safety requires stronger service assumptions. Our key observation is that the fractional peak-AoI deadline collapses exactly to an affine half-space constraint on the resource-allocation vector, turning hard real-time scheduling into time-varying constrained online convex optimization over a polyhedral safe set. A strictly causal proposal-shield-update loop enforces feasibility through one Euclidean projection per slot, the gradient step preserves no-regret behavior, and the classical virtual queue is reduced to an a-posteriori certificate. We establish closed-form static and dynamic regret bounds, a matching Omega(sqrt(T)) minimax lower bound, a margin-safe variant against execution noise, and a deadline-induced competitive ratio. On a four-sensor adversarial fluid-model trap channel, OCO-PAoI-Hard attains zero modeled-state deadline violations across all ten seeds, while four representative baselines miss between 1.65 percent and 64.0 percent of slots, and the empirical normalized regret stays below the theoretical envelope across two orders of magnitude in T.

AI 总结：该研究针对分布式边缘网络的海量数据问题，提出F²CTO方法，将分布式鲁棒核心集选择建模为带层级约束的三层优化问题，其收敛速率为O(ε^(-3/2))，经实证验证有效高效。

链接：https://arxiv.org/abs/2607.27632

作者：Yang Jiao, Kaixuan Jiao, Kai Yang, Nadjib Aitsaadi, Ilhem Fajjari, Renwei (Richard)Li

英文摘要：With the rapid advancement of the Internet of Things (IoT), massive amounts of data are generated across distributed edge networks. Training models on full data incurs significant computational overhead and storage bottlenecks, rendering coreset selection a critical paradigm. Furthermore, given the privacy-sensitive nature of local data and the escalating demand for model robustness in real-world deployments, developing an effective distributed optimization framework for robust coreset selection is vital, yet remains largely unexplored. To this end, this work first characterizes the hierarchical dependencies among coreset selection, robust optimization, and distributed learning, and formulates the distributed robust coreset selection as a trilevel optimization problem with level-wise constraints. Furthermore, to effectively solve the trilevel problem in a distributed manner, the \underline{F}ederated \underline{F}irst-order \underline{C}onstrained \underline{T}rilevel \underline{O}ptimization (F$^2$CTO) is proposed, which synergistically integrates a hierarchical composite value-function reformulation and a distributed alternating projected gradient algorithm. To the best of our knowledge, F$^2$CTO is the first method developed for distributed robust coreset selection, as well as the first distributed optimization approach for trilevel optimization problems with level-wise constraints. Additionally, we prove that the proposed method achieves a non-asymptotic convergence rate of $\mathcal{O}(\epsilon^{-3/2})$ for finding an $\epsilon$-stationary point. Extensive empirical evaluations on reliable continual learning demonstrate the effectiveness and efficiency of the proposed F$^2$CTO.

AI 总结：该研究推导了Transformer训练的有限样本泛化界，通过将训练问题转化为马尔可夫控制问题，结合量化模型与Wasserstein分布鲁棒优化，建立了Transformer泛化与分布鲁棒控制的关联。

链接：https://arxiv.org/abs/2607.27975

机构：Bilkent University（比尔肯特大学）

作者：Kağan Akman, Naci Saldi, Serdar Yüksel

英文摘要：We derive finite-sample generalization bounds for Transformers trained with dynamic programming recursions. Building on the doubly lifted, measure-valued formulation of Transformer dynamics, we view data sets as probability laws on pairs of empirical input-output measures, allowing us to interpret the training problem as a finite-horizon Markovian control problem. We then analyze a quantized model, derived by quantizing the state, action, and measure-state spaces, and derive explicit finite-sample generalization bounds using concentration inequalities for empirical laws on finite metric spaces together with a Lipschitz stability estimate for the value function. These bounds are transferred to the base model at the cost of an explicit approximation error. Finally, we show that the same machinery yields a distributionally robust control formulation of the training problem, connecting Transformer generalization to Wasserstein distributionally robust optimization.

AI 总结：TriShield是一种三层确定性防御，通过参数特征检测、有状态虚拟迭代和零效用正交投影，可完全抵御NeuroImprint攻击，且零模型效用损失、无额外通信轮次，计算开销低。

链接：https://arxiv.org/abs/2607.27940

机构：Honor Device Co., Ltd.（荣耀终端有限公司）

作者：Cheng Wei (Honor Device Co., Ltd., Shenzhen, China)

英文摘要： Federated fine-tuning of large language models (LLMs) enables collaborative training without exposing raw data. However, a recent attack, NeuroImprint [1] ( arXiv:2606.20553 ), demonstrates that a malicious parameter server can corrupt a PEFT adapter into a privacy backdoor: by assigning a dedicated memorization neuron to each training sample and ensuring each neuron updates at most once, the server can analytically reconstruct 59\%--79\% of client training data with high semantic fidelity. Existing defenses---including local differential privacy (LDP) [8] and gradient clipping---either fail against this attack or impose unacceptable utility degradation. We present \textbf{TriShield}, a three-layer deterministic defense that completely prevents NeuroImprint-style reconstruction with \textbf{zero model utility loss} and \textbf{no additional communication rounds}. TriShield consists of: (1) a \textbf{Parameter Artifact Detector} that identifies memory-neuron signatures in distributed model parameters before local training begins; (2) a \textbf{Stateful Virtual Iteration} mechanism that forces Adam/AdamW's momentum state to irreversibly entangle gradients across virtual steps, invalidating NeuroImprint's closed-form inversion; and (3) a \textbf{Zero-Utility Orthogonal Projection} operator that projects all local gradient updates onto the main-task semantic subspace computed via SVD, physically eliminating any gradient components that carry private memorization. We prove theoretically that after Layers 2 and 3, the mutual information between the uploaded gradient and any individual training sample is zero. Experiments on GPT-2 (117M) and Llama-Guard-3-1B verify that TriShield reduces NeuroImprint reconstruction rate to \textbf{0\%} across all tested attack variants, while maintaining or improving training accuracy, with less than 5\% additional GPU computation overhead.

AI 总结：提出FLAMECHE将基于元数据的CFL转化为分布式EM过程，实现加密兼容的聚类，优化了CFL三难问题的性能权衡，提升了客户端模型有效性。

链接：https://arxiv.org/abs/2607.28338

机构：UT3（图卢兹第三大学）； IRIT（信息论与电信研究所）； CNRS（法国国家科学研究中心）； INU Champollion（尚波利翁国立大学学院）； ISIS（信息系统与安全研究所）； UT2J（图卢兹第二让·饶勒斯大学）

作者：Michael Ben Ali, Imen Megdiche, André Péninou, Olivier Teste

英文摘要：Clustered Federated Learning (CFL) addresses data heterogeneity in federated settings by grouping clients with similar data distributions to enable effective training. Existing methods face a trade-off between privacy preservation, communication cost, and computational efficiency. We formalize this as the CFL trilemma, according to which improving two of these dimensions comes at the expense of the third. A prominent paradigm relies on metadata (i.e., low-dimensional representations of client datasets shared with the server) to enable communication- and computation-efficient clustering. However, such approaches are not compatible with standard FL privacy-preserving mechanisms. To address this limitation, we propose FLAMECHE, which reformulates metadata-based CFL as a distributed Expectation-Maximization (EM) procedure, restricting server updates to additive operations while preserving efficiency. This design enables compatibility with practical secure FL schemes. We conducted extensive experiments on multiple datasets under various heterogeneous scenarios. Results show that FLAMECHE improves the effectiveness of client models. It enables encryption-compatible metadata-based clustering, enhancing its positioning within the CFL trilemma.

AI 总结：针对现有归一化互信息估计器对维度敏感、数值稳定性差的问题，本文提出NMINE神经估计器，结合MINE与MI-NEE方法，在1-8维高斯数据上精度优于KSG基线，为连续多维归一化依赖度量提供新方向。

链接：https://arxiv.org/abs/2607.27710

作者：Petra Eerikinharju, Marko Tuononen, Ville Hautamäki

英文摘要：Mutual information is a general measure of statistical dependence that captures both linear and nonlinear relationships between random variables. For continuous and multidimensional variables For continuous multidimensional variables, mutual information must be estimated from samples. Because mutual information is unbounded, its values are not directly comparable across datasets, dimensions, or applications. Normalized mutual information addresses this limitation by converting mutual information into a normalized dependency score. Recent work has demonstrated the practical value of normalized mutual information in applications such as molecular dynamics { arXiv:2405.04980 } and interpretable machine learning { arXiv:2409.16768 }, but existing estimators remain sensitive to dimensionality and numerical stability { arXiv:2410.07642 }. In this paper, we propose a fully neural normalized mutual information estimator for continuous variables. The proposed approach combines a MINE-based neural mutual information estimator { arXiv:1801.04062 } with MI-NEE-inspired neural marginal entropy estimators { arXiv:1905.12957 }. Mutual information is estimated using the Donsker--Varadhan representation, while marginal entropies are estimated by learning the divergence between each marginal distribution and a uniform reference distribution, from which entropy is recovered. The resulting estimator provides a neural alternative to k-nearest-neighbor-based normalized mutual information estimation { arXiv:2405.04980 }. Experiments on Gaussian data from one to eight dimensions show that the proposed estimator improves accuracy over a KSG-based normalized mutual information baseline. These results indicate that neural estimation is a promising direction for normalized dependency measurement in continuous multidimensional settings.

AI 总结：本研究通过引入干净教师扩展信息瓶颈蒸馏框架，在CIFAR-10/100数据集上提升了干净样本分类准确率，保持对抗样本准确率，与现有双教师蒸馏方法竞争力相当。

链接：https://arxiv.org/abs/2607.27737

机构：Joint Graduate School of Mathematics for Innovation（联合创新数学研究生院）； Kyushu University（九州大学）； Faculty of Information Science and Electrical Engineering（信息科学与电气工程学院）； Université Savoie Mont Blanc（萨瓦大学）

作者：Vincent Ryusuke Takahashi, Yoshinari Takeishi, Jun'ichi Takeuchi, Kave Salamatian

英文摘要：Deep neural networks (DNNs) have achieved remarkable success in classical machine learning problems. However, they are known to be vulnerable to adversarial attacks. Countermeasures proposed in the literature, notably Information Bottleneck Distillation (IBD) introduced by Kuang et al., degrade the classification accuracy on clean inputs while improving the robustness to adversarial inputs. In this work, we extend the IBD framework by introducing an extra teacher model (clean teacher) trained with only clean inputs, into the distillation process from a robust teacher model trained by adversarial training. The features of both clean and robust teachers are transferred to the student through a cross-layer attention matrix. Experimental results on the CIFAR-10 and CIFAR-100 datasets show that the proposed method improves classification accuracy on clean samples compared to the original IBD, while maintaining similar accuracy on adversarial samples. Furthermore, our methods are competitive with state-of-the-art approaches, including the recent dual-teacher distillation framework B-MTARD, particularly in terms of the harmonic mean between clean and robust accuracy. We also analyze the impact of different training settings that have different influences on the attention module.

AI 总结：本文针对最优决策树可扩展性研究的空白，提出通用算法框架，实证分析18种搜索策略，找到的最优策略在分类任务上任何时候性能更优，回归任务运行时间提升超一个数量级。

链接：https://arxiv.org/abs/2607.28170

作者：Jacobus G. M. van der Linden, Mim van den Bos, Emir Demirović

英文摘要：Optimal decision trees (ODTs) are compact, interpretable machine learning models that globally optimize a given objective, but their scalability remains challenging. While recent work has proposed a variety of search strategies to improve scalability, the precise contribution of each strategy remains unclear. To address this gap, we introduce a general algorithmic framework for ODTs that instantiates previously used search strategies and enables the definition of new ones. This provides a common lens through which to understand and compare different strategies, which we use to empirically investigate the effect of 18 search strategies. Compared to the state of the art, the best strategy in our evaluation achieves significantly better anytime performance for classification, and improves runtime by more than an order of magnitude for regression.

AI 总结：研究针对容错量子计算中硬件噪声非平稳及分布偏移导致的性能问题，提出噪声自适应神经预解码框架QAdapt，在基准数据上显著降低量子纠错的逻辑错误率与解码延迟。

链接：https://arxiv.org/abs/2607.28422

机构：Beijing Zhongke Qhub Technology Co., Ltd.（北京中科瓴湖科技有限公司）； Center for Quantum Information, Institute for Interdisciplinary Information Sciences, Tsinghua University（清华大学交叉信息研究院量子信息中心）； Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所）

作者：Ran Miao, Rui Luo, Xiaohan Shan, Xiaoming Sun

英文摘要：Fault-tolerant quantum computing (FTQC) relies on quantum error correction to suppress physical errors and preserve logical information at scale. In practice, however, performance is constrained not only by physical noise but also by the latency of classical decoders processing rapidly generated syndrome data. This challenge is exacerbated by hardware noise that is strong, heterogeneous, and nonstationary, as well as by the simulation-to-hardware distribution shift that can substantially degrade fixed neural decoders. We present QAdapt, a noise-adaptive neural pre-decoding framework for surface-code quantum error correction. QAdapt captures local spatiotemporal correlations in syndrome data, sequentially adapts to evolving noise conditions while mitigating catastrophic forgetting, and forwards the residual syndrome to a conventional global decoder. Across 110 synthetic out-of-distribution noise configurations for rotated surface-code memory circuits, QAdapt consistently reduces the logical error rate relative to the neural pre-decoding baseline. On Google's Willow benchmark data, without target-domain fine-tuning, it achieves reductions of up to 5.79 percent in logical error rate and 9.32 percent in backend decoding latency on the residual syndrome. These results demonstrate that QAdapt provides a practical and decoder-compatible approach to improving the robustness and backend decoding efficiency of quantum error correction under evolving hardware noise.

AI 总结：该研究针对时序异质图学习中跨类型迁移与关系专业化难以兼顾、时间注入方式局限的问题，提出双分支THGFM模型，通过双路径融合与旋转时序注意力，在多学术图基准上取得显著性能提升。

链接：https://arxiv.org/abs/2607.27303

机构：RWTH Aachen University（亚琛工业大学）； Fraunhofer FIT（弗劳恩霍夫应用信息技术研究所）

作者：Yixin Peng, Diego Collarana, Er Jin, Stefan Decker

英文摘要：Temporal heterogeneous graphs offer a natural abstraction for dynamic relational systems in which diverse node and relation types co-exist and evolve over time. Learning on such graphs requires jointly modeling cross-type structural heterogeneity and the temporal dynamics of interactions, yet existing methods still struggle to reconcile parameter-efficient cross-type transfer with relation-aware specialization, and typically inject time only as additive features outside the attention kernel. We propose \textbf{THGFM}, a web-scale temporal heterogeneous graph fusion model that addresses both limitations within a unified dual-path architecture. THGFM couples a \textit{Shared-Space Temporal Attention} branch for parameter-efficient cross-type transfer with a \textit{Relational Type-Partitioned Temporal Attention} branch for relation-aware specialization, and integrates them through \textit{Dual-Path Relational--Shared Fusion}, instantiated with \textit{Type-Conditioned Non-Competitive Gated Sum Fusion}: a adaptive mechanism that assigns independent, type-conditioned feature-wise gates to the shared and specialized branches, allowing both to be amplified or suppressed without zero-sum competition. To directly incorporate relative time into the attention score, THGFM further introduces \textit{Rotary Temporal Attention}, which rotates queries and keys by half-phases of relative time before matching. THGFM consistently outperforms baseline graph transformer models on academic graphs benchmarks, delivering a $+3.25\%$ six-task mean gain, with peak relative gains of $+12.37\%$ on OAG-CS PV, $+4.87\%$ on PF-$L_2$, and $+1.18\%$ on PF-$L_1$, and $+4.24\%$, $+3.73\%$, and $+4.61\%$ on OGBN-MAG, HTAG-ArXiv, and HTAG-DBLP, respectively.

AI 总结：该研究提出首个大语言模型驱动特征自动化设计的L2O统一框架FunL2O，经多类优化任务及四种大语言模型验证，其演化特征性能优于人工设计特征。

链接：https://arxiv.org/abs/2607.27389

作者：Bingheng Li, Junyang Cai, Yupeng Zhang, Bistra Dilkina, Jayant Kalagnanam, Dzung T. Phan

英文摘要： Learning-to-optimize (L2O) methods accelerate repeated optimization by training models to predict solutions, warm starts, branching decisions, or other forms of solver guidance. A critical yet largely overlooked component of these pipelines is the feature function that maps problem instances to inputs for machine learning models. Existing L2O methods typically rely on hand-crafted features, making representation design manual and largely fixed across domains. We introduce FunL2O, the first unified framework for automating feature design through LLM-driven program evolution for L2O. In a FunSearch-style loop, an LLM proposes executable feature functions, while a fixed evaluation process retrains the original L2O model and measures downstream optimization performance. We evaluate FunL2O on linear and quadratic programming tasks involving solution prediction and warm-starting, as well as on mixed-integer optimization tasks using GNN-guided backdoor branching and Predict-and-Search. Across continuous and discrete optimization tasks and four LLMs, the evolved features consistently outperform hand-crafted representations. These results establish LLM-driven feature evolution as a general and effective approach to automating representation design in L2O.

AI 总结：针对图神经网络的过挤压问题，提出基于群论的Schreier-Coset图重连方法，可降低5%-40%有效电阻，缓解连接瓶颈并保持竞争力准确率。

链接：https://arxiv.org/abs/2607.27479

作者：Aryan Mishra, Randy Martinez, Lizhen Lin

英文摘要：The information flow in the graph neural networks (GNNs) is fundamentally constrained by over-squashing, where structural bottlenecks impede long range information propagation. Graph-rewiring methods, which modify graph topology, have been extensively used to alleviate this. However, existing approaches often introduce prohibitive structural and computational bottlenecks, fail to preserve the critical properties of original graphs, and increase the edge counts massively. We introduce a novel method Schreier-Coset Graph Rewiring, a group-theoretic rewiring method that augments the input graph with a Schreier-Coset graph derived from a special linear group. Our method provides theoretical guarantees, a graph that exhibits spectral gap and a bounded effective resistance, creating a low-resistance bypass for long-range communication. Empirical evaluations demonstrate that SCGR reduces effective resistance by 5-40% across various learning tasks, effectively mitigating connectivity bottlenecks while maintaining competitive accuracy.

AI 总结：该研究提出零样本迁移协议，通过几何重归一化缩小图训练GNN，可直接将权重迁移到全分辨率图，保留预测性能并降低训练成本，为尺度等变图架构提供了方向。

链接：https://arxiv.org/abs/2607.27767

机构：TU Delft（代尔夫特理工大学）； University of Seville（塞维利亚大学）； University of Barcelona（巴塞罗那大学）； Harvard University（哈佛大学）； ICREA（加泰罗尼亚研究与高级研究院）

作者：Robert Jankowski, Pedro Almagro-Blanco, Marián Boguñá, Melanie Weber, M. Ángeles Serrano

英文摘要：Graph neural networks (GNNs) can operate on large graphs but become infrastructure-sensitive at the scale of millions of nodes and typically require scalable training techniques for even larger graphs. This raises a central question: when can a model trained on a smaller, scaled-down replica of a graph be deployed on the full-resolution graph without retraining? We introduce a zero-shot transfer protocol in which a GNN is trained on a graph coarse-grained by geometric renormalization (GR), and the resulting weights are transferred directly to the original network. Across synthetic and real-world networks, training on GR scaled-down replicas preserves much of the original-scale predictive performance while significantly reducing training cost. We further find that learned representations and predictive trajectories remain aligned across scales. These findings suggest that structural similarity may be more important than network size in determining GNN transferability, opening a path toward scale-equivariant graph architectures.

AI 总结：本文针对循环图神经网络的过平滑问题，提出注入持续高斯噪声的方法，从理论上证明该方法可保留表示多样性，实验也验证了其有效性。

链接：https://arxiv.org/abs/2607.28185

机构：Amirkabir University of Technology (Tehran Polytechnic)（阿米尔卡比尔理工大学（德黑兰理工大学））

作者：Mostafa Haghir Chehreghani

英文摘要：Oversmoothing is a fundamental limitation of deep graph neural networks (GNNs), where repeated message passing causes node representations to become increasingly similar, eventually collapsing toward a low-dimensional subspace. This phenomenon limits the effective depth of message-passing architectures and motivates the search for mechanisms that preserve representation diversity. In this paper, we study a recurrent graph neural network in which independent Gaussian noise is injected after every propagation step and analyze the resulting architecture as a stochastic dynamical system. Under a standard global contraction assumption on the deterministic update, we prove that the hidden representations form a geometrically ergodic Markov chain admitting a unique invariant probability measure. Our main theoretical result establishes an explicit positive lower bound on the expected stationary Dirichlet energy, proportional to both the noise variance and the spectral gap of the underlying graph. Consequently, the stationary representations cannot collapse onto the constant manifold, providing a rigorous guarantee that asymptotic oversmoothing is prevented in the sense of non-vanishing Dirichlet energy. Our analysis reveals persistent stochastic perturbations as a fundamentally different mechanism for combating oversmoothing, complementing existing deterministic approaches based on residual connections, normalization, and graph rewiring. Finally, numerical experiments on both linear and nonlinear recurrent graph neural networks closely match the theoretical predictions, illustrating the emergence of a stationary distribution and the predicted dependence of the limiting Dirichlet energy on the noise intensity.

AI 总结：该研究提出TopoFormer框架，通过核心模块Topo-Scan将图拓扑结构编码为注意力友好序列，结合Transformer实现图表示学习，在相关基准上达到SOTA性能，开辟了拓扑与注意力融合的图学习新方向。

链接：https://arxiv.org/abs/2607.28259

作者：Md Joshem Uddin, Astrit Tola, Cuneyt Gurcan Akcora, Baris Coskunuzer

英文摘要： We introduce Topoformer, a lightweight and scalable framework for graph representation learning that encodes topological structure into attention-friendly sequences. At the core of our method is Topo-Scan, a novel module that decomposes a graph into a short, ordered sequence of topological tokens by slicing over node or edge filtrations. These sequences capture multi-scale structural patterns, from local motifs to global organization, and are processed by a Transformer to produce expressive graph-level embeddings. Unlike traditional persistent homology pipelines, Topo-Scan is parallelizable, avoids costly diagram computations, and integrates seamlessly with standard deep learning architectures. We provide theoretical guarantees on the stability of our topological encodings and demonstrate state-of-the-art performance across graph classification and molecular property prediction benchmarks. Our results show that Topoformer matches or exceeds strong GNN and topology-based baselines while offering predictable and efficient compute. This work opens a new path for parallelizable and unifying approaches to graph representation learning that integrate topological inductive biases into attention frameworks.

AI 总结：本研究提出基于集成共识的分子图半监督学习方法，可提升多场景下分子预测准确率，增强模型鲁棒性、降低校准误差，且单模型性能优于传统监督训练的完整集成。

链接：https://arxiv.org/abs/2607.28304

作者：Rasmus Tirsgaard, Laurits Fredsgaard, Marisa Wodrich, Mikkel Jordahn, Mikkel N. Schmidt

英文摘要：Machine learning is transforming molecular sciences by accelerating property prediction, simulation, and the discovery of new molecules and materials. Acquiring labeled data in these domains is often costly and time-consuming, whereas large collections of unlabeled molecular data are readily available. Standard semi-supervised learning methods often rely on label-preserving augmentations, which are challenging to design in the molecular domain, where minor changes can drastically alter properties. In this work, we show that semi-supervised methods that rely on an ensemble consensus can boost predictive accuracy across a diverse range of molecular datasets, task types, and graph neural network architectures. We find that training with an ensemble consensus objective increases robustness in models and exhibits an effect similar to knowledge distillation; an individual member of an ensemble trained this way outperforms a full ensemble trained in a traditional supervised fashion in almost all cases. In addition, this type of semi-supervised training reduces calibration error.

AI 总结：该研究针对GNN同图跨任务迁移的评估缺陷，提出无泄漏协议，发现迁移方向性规律，引入CTS指标，表明数据集同质性可指导机制选择以避免负迁移。

链接：https://arxiv.org/abs/2607.28525

作者：Neelam Akula, Surbhi Kumar, Murat Kantarcioglu, Baris Coskunuzer

英文摘要：Many real-world graphs support multiple predictive tasks over the same underlying structure, creating an opportunity to reuse supervision across node classification (NC) and link prediction (LP). However, existing evaluations often rely on incompatible splits, observed-graph assumptions, and negative sampling rules, making conclusions about same-graph cross-task transfer unreliable. We formalize same-graph NC-LP transfer and propose a leakage-free protocol that fixes node and edge splits, uses a shared message-passing graph that excludes evaluated edges, and employs fixed negatives for LP. Across three backbones (GCN, GraphSAGE, GPS), we find that transfer is strongly directional and predictable: NC $\to$ LP is consistently beneficial on homophilic graphs, while LP $\to$ NC is fragile and can even degrade accuracy under naive representation reuse. LP $\to$ NC becomes reliably positive mainly in a structure-dominant regime where LP is easy but NC is unsaturated, suggesting that LP acts as structural pretraining. Finally, we introduce the CoTask Score (CTS) to summarize joint NC+LP utility when a shared encoder must serve both tasks, and show that simple dataset statistics, especially homophily, can guide mechanism choice and help avoid negative transfer.

AI 总结：针对多模态持续学习中的模态贡献漂移问题，提出含基于重放和无重放版本的CMCDR方法，经实验验证其通用性与有效性。

链接：https://arxiv.org/abs/2607.27260

机构：School of Computing and Artificial Intelligence, Southwest Jiaotong University（西南交通大学计算机与人工智能学院）

作者：Zhen Zhang, Jielei Chu, Bin Liu, Tianrui Li

英文摘要：Multimodal continual learning (MMCL) aims to learn emerging knowledge from multimodal data while preserving knowledge. To mitigate forgetting, current MMCL methods usually focus on cross-modal representation alignment or semantic similarity, but they overlook whether the relative contributions of individual modalities and their interactions remain stable across incremental tasks. We term this decision-level shift Modality Contribution Drift (MCD) and quantify it with the MCD score, which combines contribution-strength and relative-reliance changes under controlled interventions on modality subsets. Theoretical and empirical analyses further explain why current MMCL methods cannot reliably mitigate this drift. To this end, we propose Continual Modality Contribution Drift Regularization (CMCDR), which preserves the modality contribution structure of previously learned tasks. Since MMCL settings differ in whether old exemplars are available, CMCDR includes both replay-based and replay-free versions. The replay-based version uses modality-subset interventions as diagnostic probes on stored old samples, compares their contribution profiles between the current model and a frozen previous model, and constrains changes in old-sample modality-specific and interaction contributions. The replay-free version uses current-task samples as probes and distills the frozen model's old-task contribution responses, thereby regularizing the observed contribution profile without exemplars. Experiments on multimodal class-incremental learning and continual visual question answering validate the generality and effectiveness of CMCDR.

AI 总结：该研究提出轻量级基础模型NEXUS，以300万参数的全连接自编码器为架构，通过对撞机数据预训练实现多领域适配，在下游任务中提升准确率且计算复杂度低于同规模Transformer。

链接：https://arxiv.org/abs/2607.27501

机构：Stanford University（斯坦福大学）； SLAC National Accelerator Laboratory（SLAC国家加速器实验室）

作者：Liangyu Wu, Qibin Liu, Alexander Yue, Julia Gonski

英文摘要：We present a lightweight approach to foundation modeling (\textbf{NEXUS}) that leverages pre-trained learning from collider physics data towards out-of-domain tasks in other scientific datasets, using a fully connected autoencoder model with approximately 3 million parameters. The model pre-trains with no supervision over a large-scale collision dataset from the Large Hadron Collider modeled by charged particle track features. Downstream tasks for collider analyses, such as kinematic regression and event classification, are developed on pre-trained model weights and achieve improved accuracy with only small labeled datasets when compared to equivalent architectures trained from scratch. The benefits of pre-training are additionally investigated through latent space interpretation and application to other domains, including gravitational waves, flood forecasting, and neural activity. Furthermore, the relative computational simplicity of NEXUS is demonstrated compared to transformer approaches at comparable scale, opening the door to power-efficient inference and real-time or edge applications of foundation models in scientific experiments.

AI 总结：本文针对表面肌电手势解码中跨会话的性能下降问题，提出电极配置无关编码器，在NinaPro DB6数据集上验证其性能优于对比方法，且特征统计对齐的无标签适配可提升所有受试者表现。

链接：https://arxiv.org/abs/2607.27568

机构：University of Saskatchewan（萨斯喀彻温大学）

作者：Jethro Odeyemi, W. J. Zhang

英文摘要：Recognition accuracy obtained during a recording session does not persist when a user puts on the electrodes again after the electrodes had previously been removed. The electrodes may have moved slightly, the skin may be drier or wetter, or the elbow may be positioned differently; these factors all contribute to day-to-day variability and therefore represent a major obstacle to implementing successful pattern-recognition based myoelectric control systems in daily practice. However, simply recalibrating a user's hand for 20 min at every doff/don event is a clearly unrealistic expectation. A montage-agnostic encoder built for cross-user, cross-montage transfer is trained here using data collected during a particular recording session, and then applied to data collected later in a different recording session without adjusting anything, on the ten intact subjects of NinaPro DB6. The performance of this approach is compared to that of a per-user LDA classification pipeline, and to that of two published approaches that only rely on source data collected from the same recording session. Carried unchanged across recording sessions, the encoder retains 0.688 macro-F1 against 0.540 for the per-user pipeline, and, on the per-window metric the published baselines use, sits above both published source-only results, a band of two points that locates the encoder rather than ranking it. Of five label-free test-time adaptations, only feature-statistic alignment improves every subject; batch-normalisation re-estimation, a standard method in the domain-adaptation literature, collapses this architecture entirely. Aligning the encoder's feature statistics to the new session recovers about what a single labelled calibration repetition would.

AI 总结：Compliance2LoRA是一种自适应超网络框架，可通过生成LoRA适配器，在单个大型推理模型上实现对任意安全策略子集的按需对齐，且不牺牲任务性能。

链接：https://arxiv.org/abs/2607.27594

机构：University of Maryland College Park（马里兰大学帕克分校）

作者：Pankayaraj Pathmanathan, Furong Huang

英文摘要：Post-training alignment in large reasoning models (LRMs) has significantly improved their adaptability to diverse safety compliance settings. However, as LRMs personalization for downstream users takes center stage, the demand for varying levels of policy compliance grows as different user-specific LRMs must adhere to distinct subsets of safety policies. Training a separate LRM for each policy subset introduces severe combinatorial overhead. While in context learning methods overcome this combinatorial overhead, they introduce additional computational challenges associated with long context generation. To address this challenge, we propose \ours, a unified adaptive hypernetwork-based framework for multi-policy compliance. In our framework, safety policies serve as customizable inputs to a LoRA adapter generator, which learns to produce policy compliant LoRA weights for downstream LRM. When added to the LRM these weights enable the generation of responses compliant with the specified policy subsets. In this work, we demonstrate that training such a hypernetwork enables on-demand policy adjustments on a single LRM without sacrificing task performance across reasoning models of different sized and different evaluation datasets. This highlights the effectiveness and practicality of adaptive hypernetwork based alignment in LRMs.

AI 总结：本文解决LoRA微调的样本复杂度问题，建立匹配的上下界，得出秩选择二分性，在合成与真实基准上验证理论预测，揭示过参数化惩罚的来源。

链接：https://arxiv.org/abs/2607.27680

作者：Arunan J

英文摘要： Low-Rank Adaptation (LoRA) has become the standard mechanism for fine-tuning large pretrained models, yet its statistical properties remain only partially understood. Existing generalization results provide upper bounds of the form O~(sqrt(rd/n)) or O~(rd/n), but a matching lower bound is missing, and the question of how to choose the LoRA rank r has no formal answer. Both gaps are closed here. A local Rademacher argument establishes an upper bound of O~(rd/n) on the excess risk of the empirical risk minimizer over rank-r LoRA, whenever the target adaptation has rank at most r. A matching minimax lower bound of Omega(rd/n) is then proved via a Fano-type packing of the rank-r subspace of R^{d x d}; the bound applies to any estimator whose output lies in the rank-r LoRA class. Combining the two yields a rank-selection dichotomy. For the constrained empirical risk minimizer, the optimal rank equals the intrinsic rank r*, and over-ranking strictly hurts. For adaptive estimators of the nuclear-norm-then-truncate type, over-ranking is harmless and the rate saturates at Theta~(r* d / n) regardless of r. Taken together, the three results characterize the statistical complexity of LoRA fine-tuning within the well-specified locally quadratic regime, and identify the empirically observed over-parameterization penalty as a property of unregularized empirical risk minimization rather than of the LoRA class itself. Predictions of the theory are verified on a synthetic trace-regression benchmark and on real LoRA fine-tuning across three (model, task) configurations covering DistilBERT and RoBERTa on SST-2 and MRPC. All configurations exhibit the predicted U-shape in validation loss, with two showing statistically significant loss inflation at large ranks (paired permutation p = 0.016).

AI 总结：该论文提出时序残差概念，指出长时序基准需将全任务成功与短阶段基线预测比较，以解释长任务失败的原因。

链接：https://arxiv.org/abs/2607.27283

机构：Tencent（腾讯）

作者：Chao Peng, Zhiheng Lyu, Peijie Dong, Hande Dong, Qiang Lin

英文摘要：Long-horizon benchmarks often show that agents fail more as tasks become longer. This observation is useful for deployment, but it does not by itself explain why failure occurs. More stages create more opportunities for ordinary errors to compound; longer tasks may also contain harder individual decisions or become harder as conversation history, tool outputs, and environment changes accumulate. We use trajectory-induced degradation to mean this last possibility: earlier execution makes later work harder. When the harmful accumulation is specifically the text visible to the model, it is often called context rot. In this position paper, we argue that to claim a "long-horizon failure", benchmarks must compare actual full-task success against a baseline prediction built from short, individual stages. We call the log-ratio between this prediction and actual success the horizon residual. The comparison must use the same agent configuration and specify in advance how stages, checkpoints, information, and budgets will be chosen. The residual shows that the full rollout differs from the chosen baseline; targeted experiments are still needed to explain why.

AI 总结：ECG-InterpBench是评估心电图基础模型可解释性的基准，采用匹配规模稀疏自编码器，在多维度对比中揭示模型可解释性差异，补充了性能导向的心电图基准。

链接：https://arxiv.org/abs/2607.27404

机构：Rice University（莱斯大学）

作者：Yixuan Duan, Wei Qiu

英文摘要：Existing benchmarks for electrocardiogram foundation models primarily evaluate downstream predictive performance, providing limited insight into whether their internal representations can be faithfully decomposed, clinically interpreted, or reproduced across independent analyses. We introduce ECG-InterpBench, a benchmark designed to systematically evaluate the interpretability of ECG foundation-model representations. ECG-InterpBench uses sparse autoencoders as standardized measurement instruments and matches their capacity across models to enable controlled comparisons. We evaluate six frozen ECG foundation models across five standardized encoder depths, five matched dictionary widths, and three random seeds, producing a 450-cell interpretability atlas comprising 75 exactly matched six-model comparison blocks. The benchmark evaluates complementary dimensions of representation interpretability, including sparse reconstruction fidelity, single-feature accessibility and coverage of 49 clinically meaningful ECG measurements, and cross-seed feature reproducibility. The evaluation further quantifies patient-sampling uncertainty, depth- and seed-dependent variation, and sensitivity to the sparsity parameterization. The benchmark reveals that ECG foundation models exhibit distinct interpretability profiles. A matched replication on MIMIC-IV-ECG confirms that reconstruction fidelity and clinical accessibility identify different leading models. The benchmark is accompanied by executable evaluation code, standardized manifests, cell-level metrics, and reproducibility audits. ECG-InterpBench complements performance-centered ECG benchmarks by providing a capacity-controlled and reproducible framework for comparing ECG foundation models across distinct dimensions of representation interpretability.

AI 总结：本研究针对现有智能体评估仅看结果的缺口，提出双评估基准ClawTrack，含320个任务及过程评分器，评估21个模型后发现其可归因推理、验证结果验证瓶颈等，助力智能体改进。

链接：https://arxiv.org/abs/2607.28037

机构：Meituan（美团）

作者：Xingjian Wu, Xuhang Zhu, Xingchen Liu, Junlin Liu, Jianing Wang, Linsen Guo, Xiaoyu Li, Xuezhi Cao, Xunliang Cai

英文摘要：As LLM-based agents are deployed in complex, multi-step workflows, a critical evaluation gap has emerged: most existing benchmarks judge only final outcomes, unable to distinguish reliable reasoning from lucky success or attribute failures to specific process deficiencies, hindering attribution in long-horizon tasks. In this work, we present ClawTrack, a dual-assessment benchmark that simultaneously measures what an agent achieves (Task Score) and how it achieves it (Process Score). ClawTrack comprises 320 tasks across 8 domains with 25+ deterministic mock services. A Process Grader scores each reasoning turn along four dimensions (goal alignment, efficiency, information utilization, and result verification), anchored by 12,541 task-specific rubric items. Evaluating 21 models over 16,000+ trials, we find that: (1) process scores effectively attribute success and failure to specific reasoning dimensions, filtering lucky passes invisible to outcome-only evaluation; (2) the four dimensions are complementary, with result verification as the systematic bottleneck; (3) the framework is robust to evaluator choice across different judge LLMs; and (4) process-based trajectory filtering yields consistent post-training improvements across model scales.

AI 总结：本研究推出整合17类超80万分子样本的Chem World化学性质预测基准，并提出Mixture-PINN物理信息神经网络框架，经实验验证其可提升预测性能，为可信赖AI系统研发奠定基础。

链接：https://arxiv.org/abs/2607.28079

机构：Cleer Science（克利尔科学公司）； Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）； Imperial College London（伦敦帝国学院）； Tsinghua University（清华大学）

作者：Tianyou Bai, Huan Wang, Mingchen Gao, Fangyue Lin, Pinze Ren, Zhenlin Zhao, Siming Dong

英文摘要：Chemical property prediction plays a critical role in accelerating scientific discovery in chemistry, materials science, and drug development. However, existing benchmarks often suffer from limited task diversity, fragmented datasets, and inconsistent evaluation protocols, making it challenging to systematically assess the reliability and generalization of AI models. In this work, we introduce Chem World, a comprehensive benchmark for chemical property prediction that integrates 17 diverse chemical datasets with over 800,000 molecular samples, covering various properties including density, electrical conductivity, solubility, and other molecular characteristics. Chem World provides a unified platform for evaluating AI models across multiple property prediction tasks. Furthermore, we propose Mixture-PINN, a physics-informed neural network based prediction framework that incorporates chemical prior knowledge into data-driven learning, improving the accuracy, robustness, and reliability of chemical property prediction. Extensive experiments on Chem World demonstrate the effectiveness of our approach compared with existing methods. By combining large-scale standardized evaluation with physics-informed learning, Chem World establishes a foundation for developing trustworthy AI systems for computational chemistry and advancing AI-driven scientific discovery.

AI 总结：针对电力系统暂态稳定评估中临界清除时间估计的难题，提出事件结构物理信息神经网络ES-PINN，实现可微临界清除边界，在IEEE节点系统实验中提升了精度与计算效率。

链接：https://arxiv.org/abs/2607.27681

作者：Baoli Hao, Chenxi Hu, Ming Zhong, Ren Wang

英文摘要：Transient-stability assessment determines whether a power system can recover after a disturbance and is therefore essential to preventing generator trips and cascading outages. A key metric is the critical clearing time (CCT), which specifies the maximum time available to clear a fault before synchronism is lost. Reliable CCT estimation is challenging because complicated fault-clearing dynamics require repeated simulations over many fault severities and clearing times. We propose an event-structured physics-informed neural network (ES-PINN) that aligns its representation with the pre-fault, fault-on, and post-clearing swing dynamics and enforces exact state chaining across event interfaces. A smooth trajectory-induced stability margin defines a differentiable approximation of the CCT boundary, enabling accurate boundary extraction, local sensitivity analysis, and optional direct CCT prediction through a distilled readout. We further prove a local residual-to-trajectory-to-CCT error estimate, in which exact event chaining eliminates separate state-interface defect terms. Experiments on IEEE 9-, 14-, and 30-bus systems show that ES-PINN consistently improves held-out trajectory and stability-boundary accuracy over matched neural-surrogate baselines across mechanical and electrical contingencies with multiple clearing configurations. Additional full-network DAE validation, multi-fault experiments, and runtime analyses further demonstrate the effectiveness and computational efficiency of the proposed framework.

AI 总结：本研究提出面向请求的计算共享（ROCS）范式，通过GLM、DCA、IKBO等技术优化推荐模型推理，在公开基准和生产 workload 中均实现效率提升且保持或改善预测质量，已大规模部署应用。

链接：https://arxiv.org/abs/2607.27744

机构：Meta AI

作者：Yuxin Chen, Liang Luo, Buyun Zhang, Jian Jiao, Boda Li, Haoyu Wang, Tongyi Tang, Ao Cai, Zijian Shen, Zhengkai Zhang, Wenyi Xie, Ryan Dick, Han Liu, Neng Shi, Bin Yu, Jianbo Xiao, Shuyao Bi, Hongtao Yu, Yuanwei Fang, Zhuoran Zhao, Sijia Chen, Yang Chen, Shuqi Yang, Qianru Li, Zikun Liu, Wei Ling, Sihan Zeng, Longhao Jin, Jiaxin Lu, Yinbin Ma, Jiawei Li, Yichen Ruan, Yong Ler Lee, Birmingham Guan, Zijian Li, Jianbo Sun, Zhengyu Zhang, Zeliang Chen, Xiaohan Wei, Yuchen Hao, GP Musumeci, Venkatesh Ranganathan, Yantao Yao, Chunqiang Tang, Wenlin Chen, Santanu Kolay, Ellie Dingqiao Wen

英文摘要：Modern recommendation models gain prediction quality by scaling feature-interaction and sequence modules, but production cost constraints cap how far systems can scale. In this work, we propose Request-Oriented Compute Sharing (ROCS), a modeling and inference paradigm that exploits a unique property of recommendation inference: each user request is evaluated against many candidates, while request-side features are shared across candidates. ROCS defers request-candidate interactions as late as possible, isolates candidate-dependent representations, and evaluates substantial portions of the model once per request rather than once per candidate, significantly improving inference efficiency while maintaining or improving prediction quality. To realize this paradigm, we develop Generalized Layer Masking (GLM) to enforce candidate isolation in feature-interaction architectures, and Deep Cross Attention (DCA) to extend request-oriented sharing to sequence architectures. To support efficient GPU deployment, we co-design In-Kernel Broadcast Optimization (IKBO) that significantly accelerates ROCS model execution. Experiments on public benchmarks show that ROCS consistently improves the quality-efficiency tradeoff across recommendation backbones. On production-scale workloads, ROCS achieves up to a 3x QPS improvement on retrieval models without quality degradation and a 0.5% relative LogLoss improvement with a 50% QPS gain on a short-form video ranking model. ROCS has been deployed across large-scale recommendation systems spanning ads and organic surfaces, retrieval and ranking stages, and more than two orders of magnitude in inference complexity, delivering significant online gains at reduced infrastructure cost.

AI 总结：针对非规则时间序列预测的现有方法缺陷，提出WrapFlow连续时间建模框架，通过连续时间分词与无模拟残差流匹配训练范式，在真实数据集上实现最优预测性能。

链接：https://arxiv.org/abs/2607.28035

作者：Tianen Shen, Zhengyu Li, Yutong Li, Xiangfei Qiu, Xingjian Wu, Bin Yang, Jilin Hu

英文摘要：Irregular multivariate time series are widely encountered in applications such as healthcare monitoring, human activity recognition, and environmental sensing. Their core challenges stem from asynchronous observations, non-uniform sampling intervals, and the fact that temporal patterns themselves carry critical dynamic information. Existing approaches either rely on discretization-based preprocessing (e.g., interpolation, imputation, or aggregation), which disrupts the underlying continuous-time semantics, or adopt continuous-time modeling via ODE-based frameworks, which typically require specialized architectures and incur substantial computational overhead due to numerical solvers. To address these limitations, we propose WrapFlow, a continuous-time modeling framework for irregular time series forecasting. On the input side, WrapFlow introduces Continuous-Time Tokenization, which directly encodes raw observation events and explicitly models long unobserved intervals via gap-aware tokens. The resulting continuous-time tokens are then processed by a standard Transformer backbone to capture long-range temporal dependencies. On the output side, we develop a simulation-free training paradigm for Residual Flow Matching, which learns conditional residual vector fields around base predictions while avoiding numerical-solver simulation and backpropagation during training. This design enables high-quality continuous forecasting using only a small number of fixed rollout steps at inference. Extensive experiments on multiple real-world datasets demonstrate that WrapFlow achieves state-of-the-art performance.

AI 总结：针对现有可解释时间序列预测器忠实性不足的问题，提出IB-Forecast框架，其通过分解预测分量与信息瓶颈优化，在保证预测精度的同时提供高忠实度解释，性能优于多种基线方法。

链接：https://arxiv.org/abs/2607.28124

作者：Xu Zheng, Wei Cheng, Zhuomin Chen, Mo Sha, Jingchao Ni, Dongsheng Luo

英文摘要：As forecasts increasingly drive decisions in fields such as energy, transportation, and healthcare, understanding the historical data behind these predictions has become as crucial as the predictions themselves. Although existing interpretable-by-design forecasters reveal their internal structures, they offer no guarantee that these structures faithfully reflect the underlying evidence driving the predictions. In contrast, while faithfulness-oriented methods explicitly verify model behavior, they are almost exclusively designed for post-hoc classification tasks. To bridge this gap, we propose IB-Forecast, an inherently interpretable multivariate time-series forecasting framework. It decomposes forecasting into a learned periodic component and a residual component computed with explainable masks over input tokens. With a budget-constrained information bottleneck, end-to-end optimization enables users to directly control explanation sparsity. With a rigorous faithfulness evaluation protocol, extensive experiments demonstrate that IB-Forecast matches the forecasting error of leading black-box models while providing faithful explanations at no additional inference cost. Furthermore, under a matched sparsity budget, these native explanations consistently surpass gradient-based, occlusion-based, and optimization-based baselines across all evaluated datasets. Ultimately, whereas the native explanations of existing interpretable forecasters exhibit poor faithfulness, IB-Forecast guarantees high explanation fidelity, requiring only 14-20% of the observations to deliver low-error predictions.

AI 总结：针对推测解码中MHA/GQA转MLA导致草稿令牌接受率降低的问题，提出功能重建方法优化转换后的MLA注意力模块，在多数任务中提升了草稿令牌接受率。

链接：https://arxiv.org/abs/2607.27269

作者：Weiye Shi, Fanxu Meng, Muhan Zhang

英文摘要：Multi-head latent attention (MLA) is increasingly important for long-context LLM inference because compact latent states replace the growing key-value (KV) cache and reduce decoding memory traffic. Yet most capable open checkpoints use multi-head or grouped-query attention (MHA/GQA), so conversion is needed to obtain MLA's cache efficiency without retraining from scratch. Speculative decoding offers complementary acceleration, but its speedup depends on agreement between draft proposals and target verification. We find that direct MHA/GQA-to-MLA conversion can sharply reduce this agreement: low-rank factorization and RoPE handling introduce attention-function errors that may be tolerable for standalone generation but substantially lower draft-token acceptance. We therefore formulate MLA draft construction as functional reconstruction rather than cache compression. Our end-to-end (E2E) method optimizes each converted MLA attention module to reproduce the post-output-projection response of its original MHA/GQA counterpart on calibration hidden states. This converter-agnostic post-conversion procedure preserves the converted cache and inference graph and requires neither verifier logits nor verifier supervision. We evaluate 192 model-converter-backend-method-task configurations spanning four Llama/Qwen draft-target pairs, TransMLA and MHA2MLA, HF and vLLM, and four 200-prompt tasks. With a 0.5-percentage-point reporting tolerance, Functional Reconstruction materially improves acceptance in 37 of 64 matched task cells, leaves 26 practically unchanged, and materially decreases one. Code and evaluation artifacts are available at this https URL.

AI 总结：本研究提出CoT-Mediate框架，结合双臂协议与来源控制干预，在VQA-RAD数据集上评估LLaVA-Med和MedGemma，发现上下文位置而非声明来源是医学VLMs使用生成推理的主要决定因素。

链接：https://arxiv.org/abs/2607.27304

机构：IEM Kolkata（印度工程管理学院 Kolkata校区）； School of UEMK Kolkata（UEMK Kolkata学院）； Heritage Institute of Technology Kolkata（加尔各答遗产技术学院）； Indian Institute of Information Technology, Kalyani（卡利亚尼印度信息技术学院）

作者：Supratik Bhowal, Subhrajyoti Basu, Aritra Gir Mahanta, Anik Pal Chowdhury

英文摘要： Medical vision-language models (VLMs) generate chain-of-thought (CoT) reasoning before answering clinical questions, but whether this reasoning causally influences predictions remains unclear. We present CoT-Mediate, a behavioral framework that perturbs a single clinically meaningful attribute within a model's own generated reasoning and measures whether the resulting prediction follows the edited reasoning. Our framework combines a dual-arm protocol comparing re-prompted evidence with prefix-forced continuation, together with a provenance-controlled intervention that varies only the attributed source of identical reasoning to disentangle reasoning mediation from sycophancy. We evaluate LLaVA-Med and MedGemma on 1,000 VQA-RAD samples each. Prefix-forced continuation consistently yields higher mediation faithfulness than re-prompting, while the provenance analysis reveals distinct model-specific deference behaviors. Across both models, removing visual evidence increases reliance on injected reasoning, whereas laterality is the least faithfully tracked clinical attribute. These results show that the mechanism used to inject reasoning substantially affects measured faithfulness and that contextual position, rather than stated provenance, is the primary determinant of whether medical VLMs use their generated reasoning.

AI 总结：研究推出3.8亿参数的EEG基础模型ZUNA1.1，其灵活性远超原模型ZUNA1，在EEG去噪与重建任务中性能达标且显著优于MNE包的球形样条插值，已以Apache 2.0许可开源发布。

链接：https://arxiv.org/abs/2607.27308

机构：Zyphra

作者：Christopher Warner, Jonas Mago, JR Huml, Beren Millidge

英文摘要：We introduce ZUNA1.1, a 380M-parameter diffusion autoencoder for flexible EEG signal reconstruction. ZUNA1.1 is capable of reconstructing variable length sequences of up to 30s, with an arbitrary number of EEG channels at arbitrary scalp locations, and can reconstruct arbitrary temporal intervals within channels in addition to reconstructing entire channels. We demonstrate that ZUNA1.1 performs at least on par with our earlier ZUNA1 model, while being far more flexible and capable of handling a wide range of reconstruction tasks. ZUNA1.1 continues to substantially outperform standard EEG denoising and reconstruction methods such as spherical spline interpolation, which is ubiquitously deployed in the MNE package. The ZUNA1.1 model is released open source under the permissive Apache 2.0 license.

AI 总结：该研究针对以太坊Sybil机器人检测的标签泄漏问题，提出泄漏感知检测方法，对比序列模型与树模型，发现XGBoost在性能、延迟及能耗上均优于Transformer序列模型。

链接：https://arxiv.org/abs/2607.27350

作者：Michał Bartnicki, Jarosław A. Chudziak

英文摘要：Sybil bots are Ethereum actors that imitate legitimate users to extract airdrop rewards or influence governance. Recent Sybil detection methods increasingly use deep learning and treat blockchain activity as a quasi-linguistic sequence. However, complex sequence models are computationally expensive for real-time monitoring, and their reported performance may be inflated by label leakage from high-signal smart contracts. We ask whether and how organic users, Sybil bots, and MEV bots differ in the structural complexity of their transaction histories; whether sequential models outperform tree-based tabular models once leakage is reduced; whether transaction order or timing provides the stronger behavioral signal; and whether the resulting models are practical for low-latency deployment. Our approach to leakage-aware Sybil bot detection consists of a Blind-Spot protocol and a Transaction Grammar representation of wallet behavior. The former eliminates shortcuts associated with high-signal contracts, whereas the latter models wallets using rhythm, EVM execution structure, and intent. We evaluate this approach on Ethereum actor classification by comparing Transformer and BiLSTM sequence models against XGBoost and SVM baselines. We contribute a framework for leakage-aware Ethereum actor classification and a Transaction Grammar representation of wallet behavior. Our results demonstrate that, under leakage-aware evaluation, XGBoost outperforms Transformer-based sequence models while providing lower latency and estimated energy use.

AI 总结：该研究提出基于压缩相似度的无训练本地发现原语，通过合成交易语法、过滤高信号合约、构建行为图，实现以太坊开放世界无需直接资金关联的 Sybil 节点发现，经多类测试验证有效性。

链接：https://arxiv.org/abs/2607.27370

机构：Faculty of Electronics and Information Technology, Warsaw University of Technology（华沙理工大学电子与信息技术学院）

作者：Michał Bartnicki, Jarosław A. Chudziak

英文摘要：Sybil attackers are Blockchain actors that adopt the characteristics of regular users to exploit airdrops or influence governance. Current methods of Sybil actor detection include constructing graphs, which requires token transfers between examined wallets. Machine learning algorithms have been employed as well, but they treat the task as a closed-set classification problem, making them vulnerable to frequent changes in attack strategies or evasion tactics. We address the following questions: can compression-based similarity differentiate Sybil bots, organic users, and arbitrage bot wallets without direct financial links? What is the effect of high-signal contracts on the discovery of Sybils, and how robust are behavioral graphs under temporal drift and adversarial perturbations? Our approach synthesizes a symbolic Transaction Grammar from EVM (Ethereum Virtual Machine) traces, capturing separately transaction rhythm, execution structure, and functional intent. The high-signal contracts are filtered with our own protocol, called the Blind-Spot Protocol. Gzip-based NCD is used to construct a behavioral graph for Sybil discovery. We validate this framework against supervised machine learning baselines, a temporal split, and synthetic camouflage stress tests. Ultimately, we contribute a leakage-aware behavioral framework for Sybil candidate discovery. Its core NCD primitive requires no supervised training and can expand suspicious seed wallets without explicit funding links. We position the method as a training-free local discovery primitive for open-world blockchain audits, rather than as a formal open-set recognition system.

AI 总结：本研究提出探索式建模（XMs），为生成式模型增添第三预训练轴，可提升多域性能与效率，还能实现端到端重构生成建模，推理步骤大幅减少。

链接：https://arxiv.org/abs/2607.27372

机构：UIUC（伊利诺伊大学厄巴纳-香槟分校）； Harvard（哈佛大学）

作者：Alexi Gladstone, Heng Ji, Yilun Du

英文摘要：The deep learning revolution, kicked off by AlexNet, taught us that end-to-end training beats decomposing a problem into hand-designed stages. Generative modeling, however, has remained the exception-despite generative models being remarkably capable, they are still not trained end-to-end. This is because, at its core, generative modeling is about handling distributions with many modes, and existing scalable approaches handle this the same way, by factoring the generation procedure, which prevents end-to-end generation. In this work, we introduce Explorative Modeling, a new paradigm that instead factors the training loop, exploring K candidate matches between model generations and data, and training on the best, so predictions commit to modes rather than blurring them. We find Explorative Models (XMs) useful in two settings. First, increasing exploration adds a third pretraining axis beyond parameters and data for existing generative models-where scaling exploration monotonically improves performance across both continuous and discrete domains (images, video, and language). Notably, gains from exploration increase with scale, climbing from 7% to 36% as data scales and from 13% to 23% as models grow, with efficiency gains more than doubling at 3x the compute. Concretely, exploration improves FLOP efficiency by 4.1x, sample efficiency by 6.2x, parameter efficiency by 47%, lifts the strongest of image-generation recipes to a near-state-of-the-art 1.43 FID on ImageNet without guidance, enables scaling how end-to-end existing models are, and unlocks scaling generalization. Second, XMs enable end-to-end reconstructive generative modeling, matching diffusion on control tasks with 16-256x fewer inference steps. Together, these results establish XMs as both a new pretraining axis for existing generative models and a standalone end-to-end generative modeling paradigm.

AI 总结：该研究发现双线性对比评判器的排序能力与其优化安全性不匹配，其存在范数漂移等问题，在多任务中无法可靠排序动作，需结合价值校准标量用于动作选择。

链接：https://arxiv.org/abs/2607.27422

机构：Stanford University（斯坦福大学）； Sesame AI（芝麻AI公司）

作者：Ayushman Singh, Siddharth Aphale

英文摘要：Good action rankings do not make a contrastive critic safe to maximize. These critics increasingly act as value-like objectives for best-of-$K$ selection, planning, and critic-guided generation. Unbounded bilinear scores can let large embedding norms inflate off-support values, but cosine bounding does not remove the failure. A controlled support decomposition attributes most raw bilinear regret to norm drift. Cosine and hybrid critics nevertheless select off-support actions from most pools and incur comparable regret. Contrastive scores are weakly calibrated or inverted in the top score decile across four OGBench navigation tasks, and they fail to order fixed-query actions by value. Bellman-trained TD-Q succeeds, including in a parameter-matched function-class control. Realized costs depend on the task: simulator rollouts reveal single-step selection costs on PointMaze and the exact-$Q^*$ toy but well-powered nulls on AntMaze and HumanoidMaze, where the controller can self-correct. A training/readout decomposition traces the lost ordering to the cosine training objective; raw-trained embeddings retain weak ordering after inference-time normalization. Candidate maximization can therefore exploit false positives caused by norm drift, score saturation, or in-support misranking. Contrastive critics remain useful compatibility rankers on navigation and manipulation tasks, but action selection requires a value-calibrated scalar.

AI 总结：针对低信噪比信道建模，该研究提出混合框架NN-CLEAN，结合神经网络与CLEAN算法，在5dB SNR下精度超96%，计算复杂度大幅降低，可作为MIMO系统实时信道估计方案。

链接：https://arxiv.org/abs/2607.27450

作者：Chaofan Deng, Linyu Sun, Jaeho Lee, Arijit Raychowdhury

英文摘要：Accurate multipath parameter estimation is critical for modern wireless communication systems, particularly in challenging low-SNR environments. Traditional Maximum Likelihood Estimation algorithms, such as CLEAN, provide high-resolution parameter extraction but suffer from prohibitive computational complexity due to exhaustive grid search. Conversely, purely data-driven deep learning approaches lack physical grounding and struggle to generalize across variable multipath densities and off-grid parameters. To address these limitations, this paper proposes Neural Network-Assisted CLEAN (NN-CLEAN), a hybrid framework that embeds a multi-head residual network directly into the iterative CLEAN extraction loop. By replacing the exhaustive grid search with rapid, parallelizable forward passes while delegating residual subtraction to exact mathematical models, NN-CLEAN isolates physical multipath parameters without accumulating non- physical errors. Extensive Monte Carlo simulations demonstrate that NN-CLEAN achieves estimation accuracy exceeding 96% at 5 dB SNR, matching the traditional Grid-Search CLEAN (GS- CLEAN) baseline, while providing a massive reduction in computational complexity and substantially outperforming subspace methods and standalone one-shot neural networks. Crucially, NN-CLEAN exhibits a near-flat scaling in execution runtime and memory consumption as batch sizes increase. This highly efficient parallelization establishes NN-CLEAN as a robust, real- time solution for channel estimation in MIMO systems.

AI 总结：本研究提出与降维（DR）方法无关的FADEx方法，通过局部线性近似等技术生成逐实例局部特征归因与失真分析，其解释鲁棒可靠，在多方面优于现有DR解释方法。

链接：https://arxiv.org/abs/2607.27463

作者：Lucas Greff Meneses, Evandro S. Ortigossa, Claudio Silva, Luis Gustavo Nonato

英文摘要： Dimensionality Reduction (DR) is a fundamental tool for high-dimensional data exploration, reducing the complexity of latent spaces of machine learning models, and assisting in the explanation of complex opaque models. However, non-linear DR techniques often function as opaque transformations themselves, making it challenging to understand how individual features influence instance positioning in the reduced space. This lack of transparency complicates the analysis and interpretation of structural patterns, hindering the ability to reason about the organization of high-dimensional data based on the projected layout. In order to address this challenge, dimensionality reduction explanation methods have shown promise in improving the understanding of the observed groups and cluster structures. Unfortunately, existing DR explanation approaches tend to suffer from limitations such as multiple attributions per feature and restricted applicability to specific dimensionality reduction methods, which hinder their use. In this work, we propose FADEx, a novel local per-instance feature attribution method that leverages local linear approximation via first-order Taylor expansion and Singular Value Decomposition to provide explanations. FADEx computes the local linear models via weighted least squares, eliminating the need for out-of-sample data mapping, making it agnostic to the DR method, while simultaneously providing local feature attributions and distortion analysis. Through qualitative and quantitative evaluations, comparisons with existing methods, and case studies, we demonstrate FADEx's effectiveness and versatility in providing explanations and analytical resources for analyzing the behavior of DR methods. The results indicate FADEx yields robust and reliable explanations, outperforming existing approaches in several aspects.

AI 总结：该研究通过对连续时间窗口分类器的对齐权重轨迹拟合HMM，从模型权重中恢复离散状态，在Fakeddit和Yelp数据集上验证了状态内迁移优势，且该优势与数据分布外的迁移结构相关。

链接：https://arxiv.org/abs/2607.27482

机构：Princeton University（普林斯顿大学）

作者：Kevin Guan

英文摘要：A temporally drifting data stream may pass through discrete regimes rather than changing continuously. We ask whether such regimes are recoverable from the weights of models trained on the stream, using a hidden Markov model (HMM) fit to the chronologically ordered trajectory of those weights. We study this question in two domains known to drift over time: multimodal misinformation detection, using the Fakeddit dataset; and sentiment analysis, using the Yelp dataset. We train classifiers on consecutive temporal windows and fit an HMM to the trajectory of their aligned weights, recovering latent states that partition each timeline into coherent phases. On both datasets, classifiers generalize better to data from windows sharing the state of their training window than to windows across state boundaries. This within-state transfer advantage survives a control for temporal proximity and modestly exceeds the advantage recovered by a naive partition into contiguous states of equal size. Although the states are estimated solely from model weights, they correlate more strongly with shifts in the data's class distribution than with the weight-space geometry used to estimate them. After class divergence and lag are residualized out, the within-state advantage exceeds its permutation null on both tasks, indicating that the states recover structure relevant to transfer beyond the data distribution. Every effect replicates on both tasks but is attenuated on Yelp, whose label distribution is more temporally stable.

AI 总结：该研究提出一种分层可解释关键实例选择模型，结合邻近树、斜树与集成学习扩展数据无关的模型，在多类型数据集上性能优于同类策略，实现了可解释性与预测效果的平衡。

链接：https://arxiv.org/abs/2607.27522

作者：Alessio Cascione, Mattia Setzu, Cristiano Landi, Paolo Maria Mancarella, Riccardo Guidotti

英文摘要：As decision-making processes grow more complex, machine learning tools have become essential for tackling business and societal challenges. However, many existing methods rely on decision-making procedures that are difficult to interpret. Since humans naturally make decisions by comparing new cases with a few representative examples, we aim to design an approach that selects such pivots to construct an interpretable predictive model. Inspired by decision trees, we propose a hierarchical, interpretable-by-design pivot selection model based on the similarity between pivots and input instances. Our method functions both as a pivot selection technique and a standalone predictive model. Extending beyond single pivots, we incorporate pairs of pivots that are used by proximity and oblique trees, as well as ensembles, which enhance the versatility and effectiveness of our proposal. Additionally, our approach is data modality-agnostic, leveraging pre-trained networks for data transformation. Experiments across diverse datasets, including tabular data, text, images, and time series, demonstrate the effectiveness of our approach, outperforming alternative instance selection strategies and achieving competitive results against state-of-the-art interpretable models while maintaining a minimal number of pivots.

AI 总结：该研究通过受控的MV-GTA模型，在BBBP和BACE数据集上验证了显式标签和性质路由的有效性，发现联合模型在多任务nDCG上优于单独专家，但未支持自由形式路由等结论。

链接：https://arxiv.org/abs/2607.27530

机构：Oakland University（奥克兰大学）

作者：Xiao Yue, Guangzhi Qu

英文摘要：Graph-text retrieval typically maps a graph and its description to a single embedding, even when a query concerns only one semantic aspect, such as a class label or molecular property. Multiple heads can separate these aspects, but a change in the query head may alter retrieval even when the wrong text is sent to that head. Such behavior demonstrates architectural channelization, not necessarily semantic routing. We examine the conditions under which this distinction can be resolved. Our controlled version of MV-GTA uses deterministic, verifiable text segments; isolated text encoders; view-specific graph heads; and relevance derived from external labels or RDKit descriptors. Correct routing and per-sample derangements form a causal test of whether retrieval depends on content. On BBBP and BACE, correct routing improves label and property nDCG by 0.305 to 0.685 over deranged training. The expected graph head exceeds the best wrong head by 0.303 to 0.453. Topology does not specialize consistently across the two datasets. In a matched three-seed comparison, one joint model obtains mean topology, label, and property nDCG of 0.720/1.000/0.877; three separately trained Single specialists obtain 0.633/0.976/0.859. Property paraphrase augmentation also improves unseen-template nDCG by 0.140 and 0.147 over a matched-exposure canonical control. Consistency and hard-template extensions, however, reduce canonical retrieval in some settings. The evidence is therefore limited to explicit, externally grounded label and property routing and observed multi-interface consolidation. It does not establish free-form routing, consistent three-view specialization, statistical equivalence to specialists, or superior downstream prediction.

AI 总结：本文研究表格基础模型的内存需求，采用模型压缩方法可实现最高7.6倍内存缩减，同时保持相近性能，降低近87%部署需求，为从业者高效部署这类模型提供见解。

链接：https://arxiv.org/abs/2607.27546

作者：Shuting Luo, Monika Mikhail Kanaan, Cameron Gordon, Anna Leontjeva, Simon Lucey

英文摘要：Tabular Foundation Models, such as TabPFN, have received a large amount of recent attention due to their performance on in-context tabular machine learning tasks, which often exceeds classical baselines. However, practical deployment considerations of these models has received less attention. In this paper we investigate the memory requirements for these models. We demonstrate that employing model compression approaches can enable memory reductions of up to 7.6 with similar levels of performance, reducing deployment requirements by nearly 87%. Our work provides insight to practitioners seeking efficient deployment of these models in practical settings.

AI 总结：本文构建多活动下里程碑驱动的初创企业随机控制模型，求解出仅依赖有效前沿曲线的最优策略，为创业决策提供了基础支撑。

链接：https://arxiv.org/abs/2607.27563

作者：Zhengli Wang

英文摘要：New venture start-ups need to ``survive'' through multiple stages of reaching milestone targets. We investigate the strategies for start-ups in a milestone-oriented setting. We examine a model of an entrepreneurial start-up firm, where its state is captured by a diffusion process. The entrepreneur can choose between multiple activities (or controls), which incur different cost and determine the drift and the variance of the process. Depending on whether the process reaches a fixed upper boundary or a lower one, the start-up firm succeeds or fails. Continuous-time stochastic models with multiple ($\ge 3$) controls are typically very challenging to deal with. In this work, we are able to completely solve for the optimal policy and provide an explicit characterization of its structure. In particular, the optimal policy only uses controls from a set characterized by a so-called efficient frontier curve that orders the controls by two intuitive measures: riskiness (drift-to-volatility ratio) and cost-effectiveness (drift-to-cost ratio). A unique feature of our model is that depending on the model parameters, the efficient frontier curves can be of different types, resulting in qualitatively different structures of the optimal policy. As far as we know, this is the first study that analyzes a stochastic control model which admits efficient frontier curves of different types. Our work provides start-up firms with intuitive measures to evaluate their activities and offers valuable insights on how the optimal strategies in a milestone-oriented setting change qualitatively contingent upon the specific scenario. We believe the results provide a foundational block in the study of entrepreneurial decision-making.

AI 总结：本文提出适用于任意电极 montage 的编码器，实现跨用户表面肌电手势识别，在 DB1、DB2 上性能优于 per-user Hudgins 线性判别分类器，训练池达一定规模后性能趋稳，自监督预训练无额外增益。

链接：https://arxiv.org/abs/2607.27565

机构：Division of Biomedical Engineering, University of Saskatchewan（萨斯喀彻温大学生物医学工程系）

作者：Jethro Odeyemi, W. J. Zhang

英文摘要：Pattern-recognition control promises a myoelectric prosthesis that responds to many intended gestures rather than one or two, but the promise has stayed in the laboratory. A recogniser trained on one person rarely transfers to the next, and useful performance usually demands a fresh round of labelled calibration from the end user. A montage-agnostic encoder is introduced that reads each electrode with shared weights and locates it by its physical coordinate rather than its index, so one architecture ingests any channel count without montage-specific parameters. Trained across users, it exceeds a per-user Hudgins and linear-discriminant classifier by 0.234 macro-F1 on DB1 for every held-out subject and by 0.108 on DB2, and falls below it on the ten-subject DB5. Each of the encoder's three key components individually accounts for more than half of its 3-shot macro F1 in an otherwise budget-matched ablation study. A controlled subject-count sweep shows the margin is close to flat from nine training subjects to thirty-nine, so the training pool binds only as a stability floor below which cross-user training fails to converge; what tracks the direction of the comparison across the three databases is instead the strength of the per-user baseline, which signal fidelity sets. Comparing against an LDA baseline depends on budget spent training models and on how good that baseline is, and self-supervised pretraining had no benefits once a supervised model was adequately trained.

AI 总结：MUGEN是一个无码本的统一运动-语言框架，通过自适应长度自编码器实现高效运动压缩，在HumanML3D和SnapMoGen数据集上的生成、检索与对齐指标均表现优异，兼顾效率与性能。

链接：https://arxiv.org/abs/2607.27581

机构：Florida State University（佛罗里达州立大学）； Texas Christian University（得克萨斯基督教大学）； University of Miami（迈阿密大学）； University of Pennsylvania（宾夕法尼亚大学）

作者：Zhankai Ye, Yukai Jin, Bingyang Wei, Bofan Li, Yusen Wu, Fangyi Li, Shangqian Gao, Xin Liu

英文摘要： Grounding human motion in language, and language in motion, is a central step toward physical AI systems that can understand, generate, and communicate human behavior. Unified motion--language systems first coupled the two directions through a shared discrete motion codebook, but quantization limits generation quality. The strongest generators buy quality back at growing cost: stacked residual codebooks enlarge the representation; masked decoding stages, long autoregressive rollouts, and denoising chains of tens to hundreds of steps stretch inference; even the continuous-latent designs among them reach their latent only through an iterative diffusion head; and none of this decoding machinery serves understanding. We therefore propose MUGEN, a unified motion--language framework that pays neither cost: no codebook, one draw. A single adaptive-length autoencoder compresses any-length motion into a few continuous latent slots, the system's only motion representation: the language model generates them for text-to-motion and reads them back for motion understanding. Depth-routed hidden states let each slot read from the transformer depth it needs, and a calibrated head predicts a joint distribution over the full latent set, so a single draw carries the text-conditional, cross-slot variation a description permits. At a decoding cost of K language-model steps, one draw, and one decoder pass, MUGEN leads language-model baselines on FID on HumanML3D while raising retrieval precision above the real-motion reference under the standard evaluator, achieves the best CIDEr and BLEU@4 scores, and surpasses the discrete-token state of the art on every retrieval and alignment metric on SnapMoGen.

AI 总结：Prox是一种两阶段无训练的SwiGLU前馈网络激活稀疏化框架，通过近似中间通道显著性构建掩码，在多类大语言模型上实现了优于基线的性能与解码加速，且兼容量化和稀疏注意力。

链接：https://arxiv.org/abs/2607.27591

作者：Jinyi Liu, Wei Chen, Pengyu Chen, Xinyi Yuan, Minghe Bai, Guoquan Wu, Jun Wei

英文摘要：Feed-forward networks (FFNs) dominate memory traffic and computation in large language model (LLM) inference, making them a primary target for activation sparsification. However, existing training-free methods suffer substantial model-quality degradation at high sparsity due to limitations in their channel-selection strategies. We observe that the SwiGLU intermediate state provides a highly effective channel-selection signal, but obtaining it requires costly dense computation. To address this, we present \emph{Prox}, a two-stage training-free framework for sparse SwiGLU FFNs. Prox hinges on the key insight: sparse execution requires only the channel mask induced by the intermediate state, which can be constructed from the magnitude ranking of its entries rather than their exact values. Specifically, Stage 1 uses input sparsity and quantized proxy weights to construct a shared mask; Stage 2 computes the selected channels exactly, enabling sparse execution of all three projections. Across ten LLMs from six model families, Prox outperforms training-free baselines at all sparsity levels, achieves up to a $1.99\times$ end-to-end decoding speedup at 70\% FFN sparsity, and is compatible with quantization and sparse attention.

AI 总结：该研究针对LLM的KV缓存内存占用问题，提出基于反事实惊奇的KV驱逐方案及快速单层近似方法，在基准测试中性能优于或可与现有最优方法竞争。

链接：https://arxiv.org/abs/2607.27600

机构：Metacognition AI（元认知人工智能）； Australian National University（澳大利亚国立大学）； Adelaide University（阿德莱德大学）

作者：Stephen Gould, Anton van den Hengel

英文摘要：Key-value (KV) cache management through compression and eviction strategies has emerged as an important research direction in recent years. Computational demands of large language models (LLMs) and their multi-modal variants during output generation can be partially alleviated by caching previous key and value calculations needed by subsequent scaled dot-product attention operations. However, this leads to another problem: the size of the resulting KV cache grows linearly with context length and quickly consumes all available GPU memory when either the prompt or the generated output are long. KV cache management periodically prunes entries from the cache thereby reducing its memory footprint while attempting to retain sufficient information for accurate generation. A by-product is faster inference speed. We propose a simple yet effective KV eviction scheme motivated by the insight that past tokens which can be well-predicted from more recent tokens are redundant and their associated keys and values can be removed from the cache. To score entries for eviction we run the model on the tokens in their original order, reusing the key and value representations already stored in the KV cache, and applying a counter-causal attention mask so that each position attends only to its future context. This is in-distribution, tied directly to the actual cache contents, and requires no additional training. To further reduce cost, we additionally propose a fast single-layer approximation that restricts the counter-causal pass to the last transformer layer, achieving a significant speedup per refresh cycle at marginal accuracy cost. We evaluate our strategy on various open-source LLMs and benchmark datasets showing competitive or improved performance over other state-of-the-art methods. Reference code is available at this https URL.

AI 总结：该研究提出\textbf{\textit{OPAL}}框架，验证决策时间信息能否为自适应实验提供正当性，其在细胞绘画数据集上实现了低误激活率，是六种方法中唯一结合非零激活与风险控制的方案。

链接：https://arxiv.org/abs/2607.27651

机构：Science and Technology Facilities Council（科学与技术设施委员会）； Diamond Light Source（钻石光源）； University of Southampton（南安普顿大学）

作者：Jia Bi, Samuel Pinilla, Chenyang Zhu

英文摘要：Adaptive laboratories choose measurements during experiments, yet most methods begin after adaptation is permitted. We introduce Opportunity-aware Policy Authorization for Laboratories (\OPAL{}), a framework that decides whether adaptation should be enabled at all. \OPAL{} uses a precommitted contract to require non-trivial adaptation, controlled target risk and positive executed value after cost. We establish an impossibility boundary: source outcomes and unlabelled target covariates cannot uniformly support non-trivial authorization under unrestricted conditional outcome shift, and derive a target-calibrated recovery. Applied to an unseen 11,265-compound Cell Painting partition, the frozen gate selected 595 compounds, captured 384 positive opportunities and achieved strictly positive executed value under least-favourable completion; its 5.18\% false-activation upper bound remained below a 7.5\% limit. Among six methods, only \OPAL{} combined non-zero activation with this risk control. Locked pharmacogenomic and finite-campaign studies distinguish policy misalignment from non-certifiability, establishing authorization as a distinct layer for safe adaptive science.

AI 总结：本文以DGCNN在SEED、SEED-IV上的实验为例，分离EEG情绪识别的评估要素，发现跨被试泛化差距非简单欠拟合导致，需区分不同评估结果的意义。

链接：https://arxiv.org/abs/2607.27655

机构：Southeast University（东南大学）

作者：Hanting Suo, Yuwen Li

英文摘要： Reported accuracy in electroencephalography (EEG) emotion recognition depends on the complete evaluation procedure, not only the classifier. We separate the target quantity, development procedure, and reporting rule, then use one archived dynamical graph convolutional neural network (DGCNN) pathway on SEED and SEED-IV as an illustrative case. In a protocol-matched subject-dependent check, the SEED result was within 1.47 percentage points of the public reference value; the 3.40-point SEED-IV difference remained unresolved. Across 30 matched SEED subject-session trajectories, checkpoint selection based on repeated test-set evaluation increased mean window accuracy from 0.7855 at epoch 80 to 0.8892. Under five-fold subject-disjoint evaluation, validation-selected checkpoints achieved training-participant trial accuracies of 0.9990 on SEED and 0.9920 on SEED-IV. Accuracy for entirely held-out participants was 0.5348 (95% conditional subject-level bias-corrected and accelerated [BCa] interval [0.4667, 0.5985]) on SEED. The SEED-IV estimate was 0.3954 ([0.3343, 0.4648]) and is reported only as secondary sensitivity evidence because its protocol-matched compatibility check remained unresolved. The observed train-to-held-out-subject gaps are inconsistent with simple optimization underfitting, but they do not isolate subject identity from implementation, preprocessing, representation, or distributional factors. Supporting analyses further showed that participant rankings depended on representation and time scale, while a development-selected tail-risk ensemble did not establish a positive gain in a separate final evaluation. Subject-dependent, subject-disjoint, and cross-session results should therefore be reported as answers to different questions.

AI 总结：本文提出以源为中心的状态演化（SCSE）循环Transformer，解决输入条件与共享循环参考的协调问题，在多个文本任务中提升了可控循环质量，验证了其设计的有效性。

链接：https://arxiv.org/abs/2607.27656

作者：Bum Jun Kim, Kohei Hayashi, Shunsuke Kamiya, Masanori Koyama, Yusuke Iwasawa, Yutaka Matsuo

英文摘要：Looped Transformers create a useful train- and test-time compute axis by reusing the same Transformer block over recurrent depth, increasing effective depth at a fixed parameter count. However, that shared block must then govern an entire trajectory of varying hidden states over trained and extrapolated depths. Furthermore, in additive-injection looped Transformers, an input-conditioned signal is reintroduced at every recurrent step, so applying the shared transition at an input-conditioned reference can still move the hidden state. In this paper, we propose Source-Centered State Evolution (SCSE), which is designed to reconcile input conditioning with reference-preserving shared recurrence. Specifically, SCSE retains input dependence through its learned anchor and initial deviation, allows nonzero deviations to drive recurrent computation while mapping zero deviation to zero, and guarantees exact anchor invariance through its zero-deviation mask. The designated anchor is thereby a one-step fixed point by construction. The zero-deviation forcing bias is the next deviation produced from the anchor itself and vanishes in SCSE, while nonzero deviations remain active and support state-dependent recurrent computation. Our theory shows that the zero-deviation forcing bias is a design degree of freedom whose task effect can be harmful, neutral, or beneficial; SCSE resolves this choice in favor of exact anchor invariance by setting the bias to zero. Across WikiText-2, WikiText-103, direct web-corpus pretraining, held-out web-text transfer, and LAMBADA completion, SCSE improves the controlled recurrent quality frontier. Ablation studies identify the learned anchor and the anchor-coordinate deviation recurrence as the primary contributors to the gain, and a trained-model case study grounds the anchor-response diagnostic in observed recurrent motion.

AI 总结：针对联邦开放世界多模态图学习的灾难性遗忘问题，本文提出FedOGL框架，通过客户端记忆保留与服务器原型共享，使性能下降降低42.67%，同时维持或提升下游任务性能。

链接：https://arxiv.org/abs/2607.27665

作者：Zekai Chen, Haodong Lu, Shihao Li, Weiwei Ji, Xunkai Li, Xun Wu, Yinlin Zhu, Rong-Hua Li

英文摘要：Federated graph learning enables collaborative training over decentralized graph data without sharing raw graph information. As such risks evolve, clients must learn emerging classes from private multimodal graph streams, retain historical categories, and reject samples outside the known class space. In this setting, clients must learn emerging classes from private multimodal graph streams while preserving historical categories and rejecting samples outside the current known class space. The core challenge is catastrophic forgetting, which in federated multimodal graphs is not merely a classifier-level failure: old knowledge can be erased through modality-semantic overwriting, topology-induced structural erosion, and federated memory fragmentation. To address this challenge, we propose \textbf{FedOGL}, a semantic-structural memory preservation framework. On the client side, FedOGL preserves historical decision behavior through replay and task-start distillation, while protecting graph-propagation memory via projection onto a globally shared structure basis. On the server side, FedOGL maintains and transfers compact category prototypes to facilitate cross-client knowledge sharing without exposing raw graph data. Extensive experiments demonstrate that, compared with the best-performing baselines, FedOGL reduces performance degradation caused by catastrophic forgetting by \textbf{42.67\%}, while maintaining or improving performance on downstream tasks.

AI 总结：该研究提出知识引导的掩码策略 VESTIGE，针对基因组Transformer微调，在古DNA重建任务中，相比标准MLM显著提升性能，且原理可适用于多种降解序列处理场景。

链接：https://arxiv.org/abs/2607.27712

作者：Angshuman Chakravertty, Rahul Maheshwari

英文摘要： Standard masked-language-model fine-tuning applies a uniform masking probability across every token position, assuming reconstruction difficulty is position-agnostic. When the degradation process is characterised and concentrated at predictable positions, this assumption fails: at peak damage sites the model can underperform a frequency-matched random predictor. We introduce VESTIGE, a parameter-free, drop-in replacement for the standard MLM collator that aligns the masking distribution with an empirically measured per-position corruption profile. We apply it to ancient DNA (aDNA) reconstruction, where cytosine deamination produces a position-dependent C-to-T / G-to-A gradient quantified per-position by mapDamage2. Rescaling so the mean C/G masking rate equals 15% - identical to standard MLM - isolates spatial redistribution as the sole variable, with model, data, seed, and hyperparameters held fixed across both DNABERT-2 runs on a mammoth CDS corpus (two specimens, seven genes). Across six terminal-zone widths and 626 paired windows, VESTIGE leads standard MLM at every width (Delta = +4.18 to +10.35 pp, all p < 10^-8), cuts validation cross-entropy by 13% (3.274 vs. 3.757), and yields ESMFold reconstructions with TM-score > 0.95 across all six reconstructions (three genes) even under damage amplified 10-30x beyond authentic PMD rates. A 1D CNN biosecurity classifier returns AUC = 0.935 and clears 98.2% of reconstructed windows, the 1.76% remainder attributable to reference-genome features, not reconstruction artefacts. The principle is domain-agnostic: any measurable position- or context-specific corruption profile - FFPE, bisulfite, metagenomic, or nanopore - substitutes directly for the PMD array, making VESTIGE a knowledge-guided training routine for intelligent systems operating on degraded or noisy sequence inputs.

AI 总结：该研究针对多通道波形相位恢复的缺陷，提出RIPPLE方法，将Griffin-Lim作为相位先验，结合校正流细化相位，在Ambisonics和地震任务中优于传统恢复方法，降低了S波极化误差。

链接：https://arxiv.org/abs/2607.27775

作者：Jaehyuk Lee, Yeajin Lee, Dayeon Shin, Donghun Lee

英文摘要：Generative models synthesize magnitude spectra with high fidelity, while phase is delegated to a recovery module---Griffin--Lim, a vocoder, or a latent decoder---applied independently to each channel. For multi-channel waveforms this delegation is costly: the physical content of spatial audio and three-component seismograms lives in the phase relationships between channels, precisely what channel-independent recovery cannot produce. The cost is also invisible, since the magnitude-based metrics common to both fields barely move when inter-channel phase coherence collapses---so a pipeline can discard the physical information in its output while still scoring well. We argue that phase should be generated, not recovered, and present RIPPLE (Rectified Inter-channel Phase with Prior-based LEarning), which reinterprets Griffin--Lim as a phase **prior** rather than a final estimator: initialized from the source phase, this prior carries the inter-channel structure to be preserved, and a rectified flow refines it toward the target under an explicit inter-channel phase loss. Tested on first-order ambisonics environment transfer and seismic cross-station translation---two physically unrelated domains---RIPPLE outperforms recovery-based pipelines on the coherence metrics that downstream analyses consume. The seismic case is decisive: across architecturally distinct generators, per-channel recovery leaves S-wave polarization error near the $57.3^\circ$ random expectation, whereas learned phase reduces it to $33.8^\circ$.

AI 总结：该研究对比序列、表格基础模型与LLM在PPM任务的表现，发现序列模型下一个活动预测最优，表格模型在时间任务具竞争力，LLM性能滞后且成本更高。

链接：https://arxiv.org/abs/2607.27797

机构：University of Mannheim（曼海姆大学）； SAP Signavio（思爱普 Signavio）； Technical University of Clausthal（克劳斯塔尔工业大学）

作者：Lennart Fertig, Lukas Kirchdorfer, Tobias Sesterhenn

英文摘要：Predictive process monitoring (PPM) leverages event logs to forecast the future of running process instances, for instance, predicting the next activity, the remaining time until case completion, or the time to the next event. While PPM research in recent years has been dominated by deep sequence models trained from scratch, such as Long Short-Term Memory (LSTM) models, foundation-model approaches---particularly large language models (LLMs)---are increasingly explored for PPM. At the same time, tabular foundation models with in-context learning capabilities offer a promising alternative but have not yet been systematically benchmarked for PPM. Thus, it remains unclear whether classical sequence-based models remain competitive in this evolving landscape. This paper compares the three modeling paradigms both conceptually and empirically through a controlled benchmark across multiple datasets and prediction tasks. The results show that sequence models consistently perform best for next activity prediction, whereas tabular foundation models are competitive on temporal tasks, with LLMs usually lagging behind despite higher cost.

AI 总结：针对直线度量下带延迟的在线聚合问题，提出学习增强型、随机型及结合两者的算法，给出相关鲁棒性、一致性或竞争比指标，经数值实验验证性能。

链接：https://arxiv.org/abs/2607.27807

作者：Tianhang Lu, Runtian Ren, Shengcai Liu, Ke Tang

英文摘要：This paper studies learning-augmented and randomized online aggregation with delays on a line metric. We consider advice given as online suggested service lengths, and evaluate the algorithms in terms of robustness and consistency. For each $\lambda \in (0,1]$, we first propose a deterministic learning-augmented \textsc{Balance} algorithm that is $(4/\lambda+1/\lambda^2)$-robust and $(4+\lambda)$-consistent. We also propose a randomized algorithm for the problem in the classical adversarial model, which is $(e+1)$-competitive against an oblivious adversary, improving over the deterministic $5$-competitive \textsc{Balance} benchmark~\cite{bienkowski2013chain}. Notably, this competitive ratio is even lower than the lower bound of $4$ for deterministic online algorithms. Moreover, we establish a lower bound of $e$ on the competitive ratio of randomized online algorithms, improving the previous lower bound of $e/(e-1)$. Besides, we combine the two ideas and obtain a randomized learning-augmented algorithm that is $(e/\lambda+1/\lambda^2)$-robust and $(e+\lambda)$-consistent. Finally, we conduct numerical experiments to complement our theoretical analysis and evaluate the empirical performance of our algorithms.

AI 总结：该研究提出对比式概念重要性（CCI）方法，将目标与对比类别间的logit边际归因于自动提取的视觉概念，可区分概念对类别对的不同效应，经ImageNet实验证实其能揭示普通概念重要性未捕捉的特定类别对模型行为。

链接：https://arxiv.org/abs/2607.27904

机构：Bielefeld University（比勒费尔德大学）

作者：Roel Visser, Isaac Roberts, Barbara Hammer

英文摘要： Concept-based explanations are a prevalent way to explain the decisions of complex black-box methods through semantically meaningful, human-interpretable concepts. To attribute the contribution of such concepts to a model's decisions, feature attribution methods are used to quantify how strongly each concept contributes to a model output. These attributions are typically computed for a single output class and therefore answer a non-contrastive "why P?" question. In many situations, however, such as cases of misclassification, class confusion, and low-margin predictions, the more natural question to ask is "why P rather than Q?". We introduce contrastive concept importance (CCI), which attributes the logit margin between a target class and a contrast, or foil, class to concepts in an automatically extracted visual concept basis. The resulting scores are signed, indicating whether a concept supports the target over the foil or the foil over the target, and can be decomposed into target-logit and foil-logit effects. This makes it possible to distinguish globally important concepts from concepts that specifically influence a class-pair distinction, including whether their effect is shared, one-sided, or directly contrastive. We evaluate the method on ImageNet class pairs using CRAFT-style concept bases, insertion and deletion curves, logit-wise decomposition analysis, and semantic class hierarchy. The results show that contrastive concept importance reveals class-pair-specific model behavior that is not captured by ordinary concept importance alone, and that highly contrastive concepts can be evaluated against semantic superclass structure to assess whether they affect fine-grained distinctions rather than broad category evidence.

AI 总结：研究针对现有世界建模机器学习范式局限于离散时间预测的问题，提出基于PT-Flow的连续时间潜在世界模型ODEWorld，解决表示崩溃问题，在视频生成和机器人控制任务中表现出色。

链接：https://arxiv.org/abs/2607.27924

机构：Institute for AI Industry Research (AIR), Tsinghua University（清华大学人工智能产业研究院（AIR））； Berkeley Artificial Intelligence Research (BAIR), University of California, Berkeley（加州大学伯克利分校伯克利人工智能研究院（BAIR））

作者：Dongxiu Liu, Haoyi Niu, Peng Cheng, Yuan Gao, Xirui Kang, Sangli Teng, Koushil Sreenath, Xianyuan Zhan

英文摘要：In the physical world we inhabit, space and time are fundamentally continuous. However, existing machine learning paradigms for world modeling are largely confined to discrete-time prediction, thereby exhibiting significant inefficiency in capturing the dynamics of physical world. We introduce Physical-Time Flow (\textbf{PT-Flow}), a novel approach that learns a continuous latent velocity field operating in physical time. Crucially, the underlying dynamics of sequential data are parameterized by an ordinary differential equation (ODE) embedded in a well-structured representation space. Under this paradigm, the prediction of future can be recast as temporal integration via an ODE solver in the compressed latent space. Building upon PT-Flow, we construct \textbf{ODEWorld}, a continuous-time latent world model that is both efficient and versatile. By extracting time-variant features and enforcing ODE properties on both the dynamical representation space and the latent velocity field, ODEWorld effectively addresses the long-standing representation collapse issue in latent world model literature. This also enables high-quality image reconstruction even after long-horizon prediction. Moreover, its continuous nature allows for arbitrary temporal resolution and even backward prediction, which is impossible for most discrete-time models. Lastly, ODEWorld can provide rich planning-oriented information to facilitate downstream policy learning. Comprehensive experiments demonstrate that ODEWorld successfully reconciles planning-conducive dynamics abstraction with visual realism, excelling in both video generation and robotic control. \href{ this https URL }{Project Website}.

AI 总结：本文提出贝叶斯域加权方法，引入Gamma先验从狄利克雷分布推断权重，实现稳定高效的域权重学习，以更少数据识别最优数据混合，适用于大规模应用。

链接：https://arxiv.org/abs/2607.27928

作者：Xiang Yuan, Kaiqing Lei, Zhenyu Jin, Jun Shu, Deyu Meng, Zongben Xu

英文摘要：The performance of Large Language Models (LLMs) is fundamentally influenced by the distributional composition of multi-domain pre-training data. While manual heuristics were prevalent in early models, they increasingly fail to capture the intricate synergies between domains as data complexity grows. To overcome the issue, a dominant approach seeks to fit a proxy function mapping between domain weights and their corresponding validation losses, and then find the optimal domain weights to minimize validation losses. These methods rely on strong structural assumptions, such as rank invariance or scaling laws, which are often violated, resulting in non-negligible estimation bias. A promising approach is to directly optimize the weighting scheme from data. However, it suffers from unstable optimization trajectory and prohibitive computational overhead, limiting its potential to search better domain weights configurations. This paper presents a Bayesian domain weighting method to infer the weights from a Dirichlet distribution via introducing Gamma prior information learned from observations. Experimental results demonstrate that proposed method could achieve stable and efficient domain weights learning, and identifies optimal mixtures while consuming substantially less data than search-based function-fitting methods, revitalizing optimization-based domain weighting for large-scale applications.

AI 总结：本研究提炼出跨领域图特征统一的四项要求，提出基于拓扑感知滑动窗口特征编码与生成式重构的图基础模型SliGFM，以实现异质节点特征的有效统一与跨领域知识迁移。

链接：https://arxiv.org/abs/2607.27966

作者：Dongxiao He, Siqi Liu, Jitao Zhao, Yawen Li, Yi Wang, Di Jin

英文摘要： Graph Foundation Models (GFMs) have recently emerged as a promising paradigm for general-purpose graph learning, aiming to learn reusable knowledge that generalizes across diverse graph domains and downstream tasks, reducing the need for specific model development. Achieving this goal requires reconciling the substantial heterogeneity in node features, graph structures, and semantic information across domains. Among them, heterogeneous node features constitute a fundamental input-level barrier, as their dimensionality and semantics vary substantially across datasets. Existing studies typically project or map heterogeneous node features into a fixed-dimensional space, often implicitly equating dimensional uniformity with effective feature unification. Yet dimensional consistency alone does not ensure that the unified features preserve informative semantics and capture transferable patterns that can support cross-domain knowledge transfer. To bridge this conceptual gap, we distill four desiderata for cross-domain graph feature unification: formal uniformity, cross-domain transferability, information preservation, and backbone compatibility. Guided by these principles, we propose SliGFM, a graph foundation model built upon topology-aware sliding-window feature encoding and generative reconstruction. SliGFM orders feature dimensions by topological smoothness and scans the reordered features with a shared sliding-window feature encoder, transforming heterogeneous features into a common space of ordered fixed-dimensional feature tokens. This formulation enables a smoothness-aware transformer to capture transferable relational patterns among feature tokens within each node, while the generative reconstruction objective encourages preservation of the original feature information.

AI 总结：针对主流张量框架表示法难读写、易出错的问题，引入通用张量运算表示法einx，简化API、统一规则，提供可与现有框架无缝集成的Python实现。

链接：https://arxiv.org/abs/2607.27987

作者：Florian Fervers, Sebastian Bullinger, Christoph Bodensteiner, Michael Arens

英文摘要：Tensor operations represent a cornerstone of modern scientific computing. However, the Numpy-like notation adopted by predominant tensor frameworks is often difficult to read and write and prone to so-called shape errors, i.a., due to following inconsistent rules across a large, complex collection of operations. Alternatives like einsum and einops have gained popularity, but are inherently restricted to few operations and lack the generality required for a universal model of tensor programming. To derive a better paradigm, we revisit vectorization as a function for transforming tensor operations, and use it to both lift lower-order operations to higher-order operations, and conceptually decompose higher-order operations to lower-order operations and their vectorization. Building on the universal nature of vectorization, we introduce einx, a universal notation for tensor operations. It uses declarative, pointful expressions that are defined by analogy with loop notation and represent the vectorization of tensor operations. The notation reduces the large APIs of existing frameworks to a small set of elementary operations, applies consistent rules across all operations, and enables a clean, readable and writable representation in code. We provide an implementation of einx that is embedded in Python and integrates seamlessly with existing tensor frameworks: this https URL

AI 总结：该研究针对开放域大语言模型训练缺乏可验证奖励的问题，提出Flux-OPD范式，利用演化上下文作为监督，通过分解反向KL目标优化，在开放域任务上性能优于现有OPD范式。

链接：https://arxiv.org/abs/2607.28022

机构：Peking University（北京大学）； Kling Team（KLING团队）； Tsinghua University（清华大学）； Shanghai Jiao Tong University（上海交通大学）； Zhongguancun Academy（中关村学院）

作者：Yuran Wang, Zekun Wang, Bohan Zeng, Ruixu Zhang, Wenxuan Liu, Liu Yang, Yifan Dai, Yang Shi, Bozhou Li, Chengzhuo Tong, Daili Hua, Yuanxing Zhang, Wentao Zhang

英文摘要：Large language model training in open-ended domains lacks verifiable rewards, making task preferences difficult to formalize as effective supervision. Contexts can convey such preferences, yet provide little additional supervision once distilled into the student, motivating contexts that evolve with student performance. However, directly using evolving contexts as in-training supervision results in an unstable distillation target and conflicting distributions, requiring mechanisms to stabilize target and downweight conflicts. In this paper, we analyze the effect of contexts through a decomposition of the reverse KL objective, revealing two findings: the student is distilled toward the geometric mean of context-conditioned teachers, and the objective contains a conflict term that measures conflicts among these teachers. Based on this decomposition, we propose Flux-OPD, an OPD paradigm that uses evolving contexts as in-training supervision to capture task preferences in open-ended domains. Flux-OPD treats the differences between context-conditioned and context-free teachers as contextual difference signals, injects them as contextual corrections into the context-free teacher anchor, and weights their correction strength using the conflict term as an indicator. Experiments on open-ended tasks show that Flux-OPD outperforms existing OPD paradigms, highlighting the potential to combine teacher supervision with evolving contexts.

AI 总结：该研究提出两阶段牛顿初始猜测策略，通过学习特征构建降维空间，结合回归模型与GMRES校正，减少牛顿迭代次数与CPU时间，加速非线性参数化PDE求解器。

链接：https://arxiv.org/abs/2607.28036

作者：Rémy Vallot (CB, Michelin), Florian de Vuyst (BMBI), Thibault Dairay (CB, Michelin), Mathilde Mougeot (CB, ENSIIE, ENS Paris Saclay)

英文摘要：It is well known that Newton's method converges faster when the initial guess is closer to a root of a system of nonlinear equations. In this paper, a two-stage Newton initial guess strategy is proposed by learning features from a parameter-space sampling and a database of precomputed solutions. The method uses discrete Newton trajectories to construct two complementary reduced spaces: a solution feature space, built from converged states, and a corrective search direction feature space, built from intermediate Newton increments. For an unseen parameter, a regression model is used to predict a surrogate solution approximation. Then, in a second step, a residual-minimizing correction is computed using a dedicated GMRES-based approach. The resulting state is then used as an initial guess for the high-fidelity Newton method, which completes convergence. The corrective step is computationally inexpensive since it only requires residual evaluations and the solution of a small least-squares problem. The methodology is weakly intrusive once the high-fidelity residual fields and a script-based programming interface are available. This strategy reduces the number of Newton iterations and decreases the overall CPU time. Numerical experiments on representative PDE problems show quantifiable speedups compared with standalone surrogate initialization. Significant speedups are observed. This generic approach can be applied to a broad class of large-scale nonlinear problems.

AI 总结：针对电商多渠道营销预算分配的预测后优化范式失效问题，提出快慢因果框架ReAlloc，经淘宝模拟与A/B测试验证可同时提升支付订单量和收入。

链接：https://arxiv.org/abs/2607.28182

机构：Peking University（北京大学）； Alibaba Group（阿里巴巴集团）； Beihang University（北京航空航天大学）； CUHK-shenzhen（香港中文大学（深圳））

作者：Changjian Liu, Tianyu Wang, Xiaoxuan Deng, WenTao Zhu, Yuwei Xu, Jungqi Jin, Yong Gao, Chuan Yu, Jian Xu, Bo Zheng

英文摘要：E-commerce platforms must allocate fixed marketing budgets across multiple channels to maximize business utility. However, standard predict-then-optimize (PTO) paradigms fail in this compositional space due to observational confounding and severe extrapolation. We formulate this challenge as a simplex-constrained uplift decision problem and propose ReAlloc, a fast-slow causal framework. Specifically, an agile Orthogonal Teacher extracts unbiased local gradients from short-term logs, while an Explanation-Guided Student distills them into a structured marginal field over long-term horizons. This design enables support-aware, conservative decisions that capture cross-channel substitutions. Extensive simulations and large-scale online A/B tests on Taobao platform demonstrate that ReAlloc achieves simultaneous lifts in both pay order and income.

AI 总结：本文提出Gap指数，一种度量降维散点图空白区域失真的质量指标，对高视觉影响的小结构变形敏感，计算快速且可解释。

链接：https://arxiv.org/abs/2607.28324

机构：Eindhoven University of Technology（埃因霍温理工大学）

作者：Jaume Ros, Alessio Arleo, Fernando Paulovich

英文摘要：Quality metrics play a crucial role in the proper use of dimensionality reduction projections for visual analysis of high-dimensional data. They quantify the degree of distortion of a projection compared to the high-dimensional data and provide a reliable indication of how confident users can be in the structures they see in the resulting layouts. However, most popular metrics focus on capturing direct relationships between points (e.g., distances or neighborhoods) while neglecting distortions in empty areas of the layout, even though these often compose visually relevant features of a 2D layout. In this paper, we introduce the Gap Index (GI), a quality metric for 2D projections that captures visual distortion by measuring spatial distortion in empty areas of a projection. It does so by decomposing the space into empty triangles, which are then compared to their high-dimensional counterparts to compute the deformation. This per-triangle deformation can be aggregated into a single scalar value or overlaid on a projection to visualize regional distortion patterns. Results show that, contrary to popular quality metrics, the GI is sensitive to small structural deformations that have high visual impact. It is also fast to compute and interpretable.

AI 总结：该研究提出LedgerMind，通过结构化证据账本及三层依据协议等组件，解决多模态智能体推理中最终答案准确率无法反映轨迹可信度的问题，在多模态基准上同时提升了答案准确率与轨迹可信度。

链接：https://arxiv.org/abs/2607.28374

机构：The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））； The University of Hong Kong（香港大学）； Tsinghua University（清华大学）； University of Sussex（萨塞克斯大学）

作者：Enjun Du, Hange Zhou, Chenxu Du, Siyi Liu, Zirong Chen, Ziyu Zheng, Yongqi Zhang

英文摘要：Multimodal agents for visual question answering increasingly operate as multi-step trajectories that interleave perception, retrieval, and reasoning, yet evaluation still largely reduces to final-answer accuracy. This aggregate signal cannot tell whether a correct answer was reached through grounded evidence, language priors, or accidental error cancellation. We propose to treat a multimodal agent trajectory as a provenance-constrained state machine: tool outputs are normalized into a Structured Evidence Ledger that serves as the trajectory state, downstream reasoning and decision claims may cite only active ledger entries, grounding is checked at the entity and numeric level, and repair is realized as typed state transitions that cannot introduce content without tool-produced provenance. We instantiate this design as LedgerMind (Provenance-Constrained Multimodal Agentic Reasoning with a Structured Evidence Ledger), augmented by a Three-Layer Grounding Protocol, an Adaptive Dual-Path Dispatcher that matches reasoning depth to question complexity, and an Event-Triggered Verification-and-Repair engine with a formal provenance non-amplification guarantee. We use LedgerMind to target four recurring failure patterns that final-answer accuracy tends to obscure: unsupported intermediate reasoning, citation-backed entity hallucination (Phantom Grounding), over-reasoning on simple queries, and repair-time amplification. Experiments across multiple multimodal reasoning benchmarks and backbone MLLMs show that LedgerMind improves both answer accuracy and trajectory-level faithfulness.

AI 总结：针对GUI智能体因决策时间关键路径的自回归解码延迟导致瞬态事件失败的问题，提出AAPT策略树方法，提升了决策窗口内的动作成功率，验证了分支路由为关键瓶颈。

链接：https://arxiv.org/abs/2607.28399

作者：Zihan Dong, Rui Qian, Qishi Zhan, Dongshen Peng, Kaixin Li, Yu Li

英文摘要： Computer-use agents often fail on transient GUI events because they produce the correct action only after the relevant window has already closed. We identify the main cause as expensive autoregressive decoding on the decision-time critical path. We propose Adaptive Anticipatory Policy Trees (AAPT), which eliminates this delay without modifying the underlying model. During idle screen periods, the same frozen multimodal model constructs a bounded conditional policy tree with observable guards, pre-authorized actions, and branch-specific deadlines. The tree is sized to cover the model's own decoding latency. When an event occurs, a lightweight observer matches change-gated frames to a prepared branch and immediately executes the corresponding action without generating new text. In paired trials with pre-registered endpoints and exact McNemar tests, AAPT improves the success rate from 0.50 to 0.79 within a contested decision window ($p=1.8\times10^{-3}$), while producing no incorrect actions. Both open-loop and predict-and-replan baselines achieve zero success because they still decode during execution. A preparation-time sweep shows that the gain emerges where the latency-based tree-sizing rule predicts, and ablations reveal three key requirements: fast observer decoding, valid tree planning, and accurate branch routing. A pre-registered oracle probe rejects our initial hypothesis and instead points to branch routing as the causal bottleneck. We further reproduce the effect on an independent general-purpose multimodal model over 126 paired trials ($p=4.9\times10^{-13}$). On an external benchmark, AAPT matches the overall performance of a reactive baseline, although the two methods exhibit complementary strengths. Together, these results suggest that AAPT performs best when candidate actions can be enumerated in advance, whereas reactive execution remains stronger when they cannot.

AI 总结：该研究针对潜在世界模型正则化中EP目标函数对尾部样本校正梯度不足的问题，提出QQWorld方法，通过分位数-分位数匹配目标函数及跨批次QQ技术，提升了LeWM在四个控制环境中的规划成功率与高斯对齐效果。

链接：https://arxiv.org/abs/2607.28415

作者：Zhoushun Yu, Xiaoyu Hu, Xiangyu Xu

英文摘要：Latent world models enable efficient planning by predicting future states in a compact representation space, but their performance depends critically on the quality of the learned latent distribution. LeWorldModel (LeWM) regularizes its latents toward an isotropic Gaussian using the Epps-Pulley (EP) objective. We show that the corrective gradients of EP rapidly vanish for isolated tail samples, leaving heavy-tailed deviations insufficiently controlled. To address this limitation, we propose QQWorld, which replaces EP with a quantile-quantile matching objective that directly aligns projected latent samples with rank-matched Gaussian quantiles, thereby maintaining effective corrective gradients in the tails. We further develop cross-batch QQ, which enlarges the effective ranking pool using detached samples from previous batches, and characterize its bias-variance trade-off. Across four control environments, QQWorld effectively improves the average planning success rate of LeWM, while consistently yielding better Gaussian alignment and thinner latent tails.

AI 总结：该研究提出短期图记忆模块，用于在有限神谕预算下优化分子，可提升生成器的前10名分数，且适用于多款生成器，为合理分配神谕预算提供了方法。

链接：https://arxiv.org/abs/2607.28437

机构：Stony Brook University（石溪大学）； Novo Nordisk（诺和诺德公司）

作者：Jiannan Yang, Veronika Thost, Xiang Ling, Tengfei Ma

英文摘要：Molecular optimization is commonly performed under a limited oracle budget, which makes deciding what to evaluate as important as deciding what to generate. We introduce short-term graph memory, a plug-in module that preserves the generator architecture and native update rule while learning from previously evaluated molecules to prioritize subsequent oracle queries. The module maintains an online graph neural surrogate that pre-screens each round's candidate pool, so the fixed oracle budget is spent on molecules with higher predicted utility. Applied to a fragment-based generator on a standard molecular optimization benchmark, it improves the mean top-10 score at no extra oracle cost and never falls behind the base on any oracle; the gain extends to all four generators we tested at a tight budget of one thousand calls. We then analyze how surrogate-guided selection interacts with the exploration and exploitation behavior of different generators. Its benefit at larger budgets is consistent with two properties of the backbone: how broadly it searches, and how effectively its native search already exploits oracle feedback. We provide a simple way to spend a fixed oracle budget more selectively, and evidence on which generators benefit from it.

AI 总结：该研究基于Qwen2.5系统，通过实验证实阶段重放分歧由KV缓存携带，固定前缀精度会影响分歧表现，双向缓存移植可复现分歧轨迹，精确令牌重放无需保留实时状态保真度。

链接：https://arxiv.org/abs/2607.28495

机构：Openhagen（奥彭哈根）

作者：Alexander Boesgaard Lorup

英文摘要：Stage-replay diagnostics reconstruct intermediate token prefixes and treat fresh-prefill continuation as continuation from the decoder state that originally reached the prefix. We audit that assumption at a whole reasoning-stage boundary in a Qwen2.5-derived system. A matched 200-item experiment compares retained live cache with one-shot prefill of identical integer tokens and places an exact replica on both sides. In BF16, replicas remain exact while the constructions differ on 166 suffixes and 20 correctness labels; the accuracy difference is only one point (paired 95% CI [-3.5, +5.5]). A fixed-prefix 2x2 holds all 200 token states constant while crossing construction and precision. The BF16 disagreements recur, whereas FP32 produces no decoded disagreement (95% Wilson upper bound 1.88%). A prospective bridge makes token-by-token incremental and retained live caches bit-exact on 12/12 rows; an all-200 saved-ledger audit reproduces every retained trajectory and comparison fingerprint. Bidirectional transplantation of all 48 key/value layers makes every tested divergent continuation follow its cache donor, both on a selected set at the primary checkpoint (24/24) and an outcome-blind replication at a later checkpoint (43/43). Exact-token replay can therefore be repeatable without preserving live-state fidelity. On the tested states, boundary K/V cache is a causally sufficient carrier of the divergent trajectory, while numerical precision moderates its behavioral expression.

AI 总结：KAISEN是一个五阶段临床风险模型子群体公平性审计流程，经合成基准测试揭示了审计各环节的特性与局限性，相关复现资源已公开。

链接：https://arxiv.org/abs/2607.28608

机构：Massachusetts Institute of Technology（麻省理工学院）； Hopewell Valley Central High School（霍普韦尔谷中央高中）； East Brunswick High School（东布伦瑞克高中）

作者：Sparsh Roy, Samuel Girmachew, Nishita Chavan

英文摘要： Clinical risk models routinely achieve strong aggregate performance while producing materially different error rates across patient subgroups. Audit pipelines have been proposed to catch this, but their components are rarely stress-tested, so it is unclear which parts of an audit can be trusted and under what conditions. We present KAISEN, a five-phase audit pipeline covering subgroup stratification, disparity measurement, mechanism diagnostics, post-hoc mitigation, and drift monitoring, evaluated to the point of failure on a synthetic benchmark of 16 disease tasks, 15 social-determinant axes from Healthy People 2030, and three prespecified intersections. Four findings follow. (i) Significance tracks each axis's gap against its own minimum detectable effect: rank correlation between significance count and raw equalized-odds difference (EOD) across the 15 axes is rho = 0.56, rising to rho = 0.78 once EOD is standardized by that floor. (ii) Per-group threshold optimization reduces EOD in 48 of 48 held-out runs (paired delta = -0.285, 95% CI [-0.313, -0.252]), while group-wise Platt scaling -- the better calibrator -- behaves as a coin flip on EOD (19 of 48 runs improved, 95% CI [0.26, 0.55]) with mean effect near zero, so what an audit should report is the variance, not the average. (iii) The mechanism diagnostic classifies 144 of 144 controlled cases correctly but recovers none of 48 model-driven cases under proxy misspecification, with no signal that it failed. (iv) CUSUM failures and false alarms track cohort realization far more than disease: at the reference threshold, all 27 false alarms and 7 of 8 missed shifts come from different seeds (chi-squared p = 0.002), so a threshold tuned on one cohort fails to transfer. All results are synthetic with known ground truth and do not establish clinical validity. Code, artifacts, and scripts reproducing every number are released.