机器学习学术速递[9.24]

AI 总结：本研究将Muon谱更新几何应用于局部学习，提出漂移契约步长规则，显著提升深度鲁棒性并简化超参数调整，在CIFAR-10上优于局部Adam，且证明优势源于谱几何本身。

链接：https://arxiv.org/abs/2609.26811

作者：Fabien Polly

英文摘要：Local learning trains each layer with its own auxiliary loss and no global backward pass, which makes layer updates structurally parallel. Two problems have kept it marginal: accuracy degrades as depth grows, and hyperparameters are fragile. We apply Muon-style spectral update geometry (momentum orthogonalization with spectral step scaling) to per-layer local updates, an intersection not previously studied. On CIFAR-10 MLP benchmarks with local linear heads, a single step-size setting is the best value in our tested grids from width 128 to 2048 and from depth 12 to 48, while local Adam requires re-tuning along both axes and still collapses at depth 48 (31.3 percent re-tuned per depth, 19 percent with its depth-12 setting transferred, vs 42.7 percent for the spectral update at its unchanged setting). At five seeds and width 512 the spectral update leads local Adam by a clear margin (48.9 +/- 0.5 vs 46.6 +/- 0.3). Prospectively specified controls attribute the transfer and most of the depth robustness to the spectral geometry itself rather than to any step-size rule on top of it. We additionally formulate the step size as a drift contract, lr = epsilon / RMS(input), which bounds each layer's weight-induced pre-activation change per step, conditioned on its current input. The contract yields a small gain over the best fixed learning rate where that baseline is measured, makes the step size interpretable, and provides a per-layer, input-conditioned drift bound that standard optimizers do not offer. We report one negative result: with RMSNorm and weight decay in the trunk, the stability benefit of spectral updates accrues to global rather than local training, so the local advantage concentrates precisely where normalization is absent.

AI 总结：QUARTET通过因果随机游走采样和四分支交叉注意力，在局部与全局层面增强图Transformer，于RelBench v1分类任务中匹配或超越现有最先进模型。

链接：https://arxiv.org/abs/2609.26855

机构：Capital One（第一资本）

作者：Kyaw Hpone Myint, Nan Jiang, Xiang Li, Zhe Wu, Alexandre G.R. Day, Pranab Mohanty, Giri Iyengar

英文摘要：Relational Deep Learning (RDL) models multi-table databases as heterogeneous temporal graphs, and graph transformers currently achieve state-of-the-art performance on benchmarks like RelBench. However, the current leading model, RelGT, suffers from two key limitations: its random local sampler yields loosely connected subgraphs that hinder message passing, and its global attention module relies on a single, seed-feature-based memory that ignores broader macro-level dynamics. To overcome these limitations, we introduce QUARTET, an expressive graph transformer architecture that applies full self-attention on local subgraphs while enriching global context through cross-attention branches. Specifically, QUARTET employs a Causal Random Walk (CRW) sampler based on recency-truncated Personalized PageRank (PPR) to extract compact, hub-robust, and densely connected local subgraphs without temporal leakage. Concurrently, a quad-branch cross-attention module integrates global context from four complementary perspectives: seed feature, seed topology, temporal dynamics, and collaborative dynamics. Across the RelBench v1 classification tasks, QUARTET consistently matches or outperforms the current state-of-the-art graph transformer baselines (HGT and RelGT). Ablation studies confirm that the CRW sampler significantly enriches local neighborhood quality, while the global branches provide essential, task-specific predictive gains.

AI 总结：本文提出一种全协方差平滑方法，通过交叉协方差恒等式实现贝叶斯神经网络在线自适应，在非平稳分类、在线动力学学习和视觉-语言-动作模型策略自适应中优于现有平滑方法。

链接：https://arxiv.org/abs/2609.27244

机构：Carnegie Mellon University（卡内基梅隆大学）； Software Engineering Institute, Carnegie Mellon University（卡内基梅隆大学软件工程研究所）； Fujitsu Research of America（富士通美国研究院）

作者：Oren Wright, Haoming Jing, Qiaoan Shen, Koichiro Niinuma, Yorie Nakahira, José M. F. Moura

英文摘要： A neural network's layers can be treated as time steps of a state-space model, turning Bayesian training into a smoothing problem: a forward pass propagates Gaussian moments through the network, and a backward Rauch--Tung--Striebel pass updates the weight posteriors in closed form. Such methods learn from each observation in a single pass, in an uncertainty-aware manner, and without gradient-based iterations or replay, which makes them well suited for online adaptation and data-efficient learning. Existing smoothing-based methods, however, are restricted to diagonal covariances across activations, discarding correlations between neurons. We overcome this limitation via a cross-covariance identity that enables full-covariance propagation through a network's nonlinear activations. We derive a one-step-per-layer smoother that approximates as Gaussian only each layer's affine output, and that applies both to deterministic systems with noisy observations and to stochastic systems described by output statistics. We demonstrate this method in non-stationary classification, online dynamics learning, and policy adaptation of a vision-language-action model, and find that it is generally more accurate than other smoothing-based methods.

AI 总结：提出神经发生网络（NGN），通过可微边界学习有序结构组件的数量，实现结构容量作为可微计数优化，并在多种架构上验证了与固定规模模型相当的性能。

链接：https://arxiv.org/abs/2609.27291

机构：Cornell University（康奈尔大学）

作者：Lixing Li

英文摘要：Neural network size is usually chosen before training, separating architecture selection from weight optimization. We introduce the Neurogenesis Network (NGN), a differentiable parameterization for learning how many ordered structural components a model should use. For each ordered component group, one learnable boundary selects an active prefix while the model parameters are trained. The boundary can grow from a compact initialization and can be deployed by discarding components beyond the learned boundary. Controlled experiments examine convergence of the learned boundary, the performance of deployed prefixes, and comparisons with fixed-size models and alternative approaches to learning capacity. We then apply the same mechanism to MLPs, convolutional and graph networks, Transformers, state-space models, LoRA, and adapters. Across these settings, deploying only the learned prefix usually changes performance little, and the selected architectures perform similarly to fixed models trained at the same size. These results show that structural capacity can be optimized directly as a count.

AI 总结：KITE通过KV不变扩展实现模型从小到大的高效训练，并降低推理成本；SST双塔解码器在可比计算下优于更大基线模型。

链接：https://arxiv.org/abs/2609.27294

机构：StepFun

作者：Zhiheng Hu, Yixun Wei, Jian Zhou, Yizhuang Zhou, Ji Li, Xing Chen, Yang Li, Bojun Wang, Yibo Zhu, Xiangyu Zhang, Daxin Jiang

英文摘要：Scaling a language model is not only a question of final quality: the architectural choice determines how much computation is spent during training, prompt processing, and autoregressive decoding to achieve certain model quality. An ideal model architecture should lower all above computation costs to facilitate scaling to a larger model, while ensure the larger model indeed outperforms smaller baselines. We introduce KV-Invariant Transformer Expansion (KITE), a scaling paradigm that achieves this goal. It trains the model from a smaller size to a larger size (i.e., saving training costs via upcycling), while places newly added parameters in regions that do not affect attention KV. Consequently, during inference, prefilling KV only relies on the smaller part of the model, so the inference costs are saved. As a concrete instantiation, we present Step Scale Transformer (SST), a two-tower decoder in which one tower produces KV and the other reads them. At comparable cumulative training compute, SST, a 67B MoE model with 2.15B active body parameters per decode token, achieves lower training loss than 47B and 63B MoE Transformers with 1.48B and 2.02B active body parameters, respectively, while reducing estimated inference cost by 6.7% and 31.6%.

AI 总结：针对振动诊断中标签稀缺和边缘设备计算受限的问题，提出紧凑振荡状态空间模型DualRes，融合振动互补频谱视图，以选择性振荡记忆处理时间模式，仅含39,528参数，在齿轮箱基准上多数标签预算下达到最先进性能，显著提升宏F1并减少存储。

链接：https://arxiv.org/abs/2609.27411

机构：Georgia Institute of Technology（佐治亚理工学院）

作者：Mainak Mallick, Seung-Kyum Choi

英文摘要：Machine fault diagnosis from vibration requires learning from scarce labelled fault recordings while meeting the computational constraints of edge devices for local inference. We introduce DualRes, a compact oscillatory state-space model that combines two complementary spectral views of vibration, capturing rapid changes and fine frequency structure. Time-aligned views are processed by selective oscillatory memory, which learns how long to retain temporal patterns. The encoder contains 39,528 parameters. We evaluate supervised learning across six bearing datasets and a gearbox benchmark, with an additional gearbox pilot. Recording-level splits and explicit accounting of labelled duration distinguish data efficiency from repeated exposure to correlated samples. On the main gearbox benchmark, DualRes achieves state-of-the-art performance among the nine evaluated methods at six of seven label budgets. With about six labelled seconds per class, it improves macro-F1 by 16.1 percentage points over the next strongest comparator. On the same benchmark, DualRes achieves a 1.44-fold recording-level speedup and a 24.8-fold reduction in checkpoint storage relative to a selective state-space baseline under matched hardware and runtime conditions. Bearing results reveal task-dependent trade-offs. These findings support oscillatory memory as a compact approach to vibration diagnosis under limited labelled exposure.

AI 总结：本研究提出半监督框架，利用高保真特征工程（如KeyLinker和SSU）在1.63亿笔交易中检测非法比特币流动，证明数据质量优于数量，F1达0.84。

链接：https://arxiv.org/abs/2609.27936

机构：Skolkovo Institute of Science and Technology（斯科尔科沃科学技术研究院）； Moscow Institute of Physics and Technology（莫斯科物理技术学院）

作者：Yekaterina Smolenkova, Nickolay Larionov, Nikolay Ivanov, Yury Yanovich

英文摘要： Detecting illicit cryptocurrency transactions is hampered by extreme class imbalance, adversarial obfuscation, and a scarcity of reliable labels. While semi-supervised learning (SSL) offers a promising solution by leveraging unlabeled data, we show that its success is not guaranteed by data volume alone but is contingent on data quality. We introduce an SSL framework for detecting illicit Bitcoin flows in Shared Send Mixers (SSM) transactions, built on a comprehensive historical dataset comprising 163 million transactions. Our main conclusion is that the success of SSL depends on data quality rather than volume: high-fidelity features such as KeyLinker address clustering and Shared Send Untangling (SSU) complexity metrics achieve an F1 score of 0.84 on unlabeled data. Finally, we empirically show that common heuristics like One-Time Change (OTC), though abundant, introduce noise, while strategic reliance on higher-fidelity features like KeyLinker is essential. Our work establishes that in blockchain forensics, the path to better performance lies in smarter feature engineering for data quality, not just larger datasets.

AI 总结：本文提出精确分位数均衡（EQB）与负载误差注入（LEI），分别改进MoE训练的全局与局部负载均衡，在7.5B参数模型上提升性能并优于GShard损失。

链接：https://arxiv.org/abs/2609.28053

机构：Aleph Alpha

作者：Pit Neitemeier, Jiaze Li, Alessio Serra, Philipp Scholl, Sohir Maskey

英文摘要：Mixture-of-Experts (MoE) training requires global load balance to prevent expert under-utilization and local balance for efficient expert-parallel execution. Existing distributed Quantile Balancing (QB) uses shard-dependent or approximate global quantiles, while token-independent expert biases cannot ensure microbatch-level balance. We introduce Exact Quantile Balancing (EQB), which computes exact global-batch BF16 quantiles with negligible communication, and Load-Error Injection (LEI), which injects local load errors directly into router-score gradients. On 7.5B-parameter MoEs trained for up to 500B tokens, EQB improves global balance and downstream performance over naive QB, while LEI improves local balance and outperforms the GShard loss at comparable quality.

AI 总结：PhyMo提出物理场模态，通过PDE关联算子组织异构测量，经三阶段学习在五个数据集上超越最强基线，实现AI4Physics多模态表示学习的最优性能。

链接：https://arxiv.org/abs/2609.27554

作者：Henan Sun, Haitao Hu, Jin Liu, Jianfeng Zhang, Lujia Pan, Nuo Chen, Jia Li

英文摘要：Multimodal learning is emerging as a powerful paradigm for AI for Physics (AI4Physics), where predicting physical systems requires the joint interpretation of heterogeneous observations, measurements, and domain knowledge. However, existing approaches typically represent physical quantities and governing equations as generic numerical or textual tokens, overlooking the physical constraints that determine their spatiotemporal interactions. To address this limitation, we introduce the \textbf{physical-field modality} and propose \textbf{PhyMo}, a physics-grounded multimodal framework that organizes heterogeneous measurements through PDE-associated operators. PhyMo follows a three-stage learning procedure: the physical-field encoder is first pretrained through field reconstruction under PDE residual supervision, its representations are subsequently aligned with visual embeddings in a shared latent space, and the fused multimodal representations are finally processed by corresponding downstream prediction heads. Experiments on five datasets spanning diverse physical environments show that PhyMo achieves state-of-the-art performance, compared to the strongest baseline on each dataset, demonstrating the superiority of PhyMo on multimodal representation learning in AI4Physics.

AI 总结：本研究利用260万份期权合约比较机器学习方法检测Black-Scholes模型偏差的能力，发现保留领域结构的树集成方法优于抽象嵌入，且偏差反映市场结构而非模型伪影。

链接：https://arxiv.org/abs/2609.27764

机构：Stanford University（斯坦福大学）

作者：Juli Huang, Jake Cheng, Rupert Lu

英文摘要：We treat options pricing as a representation problem: can machine learning detect systematic deviations from Black-Scholes using 2.6M real option contracts? We compare three regimes: learned abstract embeddings (Kernel PCA), preserved domain structure (tree-based ensembles), and neural network validation. Tree-based methods outperform kernel dimensionality reduction by 21.5 percentage points (93.8% vs 72.3%), and domain-expert features (Greeks, moneyness) outperform engineered features. NN-based and BS-based deviation labels agree 99.9974% of the time, suggesting deviations reflect market structure rather than model artifact. We conclude that in domains with expert-designed symbolic features, preserving structure beats learning abstractions. We make no claim of exploitable mispricings.

AI 总结：提出可迁移证据重建（TER）自监督方法，通过跨组读取器测试学习结构化证据解码规则，在连续血糖监测14任务上显著提升预测性能。

链接：https://arxiv.org/abs/2609.28199

作者：Tian Zhou, Bingqing Peng, Linxiao Yang, Wenwei Wang, Mengni Ye, Beverly Jin, Zuyi Zhu, Jinjie Gu, Liang Sun

英文摘要： Long physiological recordings contain many routine measurements, while predictive information is often concentrated in rare events, sustained burden, and recurring temporal patterns. Masked autoencoding recovers measurements; contrastive learning aligns views. We study self-supervision that explicitly prioritizes structured signal evidence. We introduce transferable evidence reconstruction (TER), which constructs evidence from unlabeled recordings, fits a fresh low-capacity reader on one recording group, and requires that reader to recover the same evidence in another group without refitting. Differentiating through this cross-group test learns representations with transferable evidence-decoding rules; the evidence guides self-supervision but is not used as a downstream feature. For continuous glucose monitoring (CGM), an observation-aware daily encoder and clock-aware multi-day memory bind glucose level and change to recorded time while organizing up to seven days of history. On the 14-task leaderboard, TER improves the strongest prior overall PR-AUC/ROC-AUC/Macro-F1 scores by 5.51/4.43/2.80 percentage points and sets a new best metric on 12/14 tasks. These leaderboard gains are 2.0-2.9 times the respective gaps between the two strongest baselines. With public pretraining data, folds, and the linear probe matched, TER outperforms our GlucoFM reproduction by 6.09/5.52/2.72 points. Target-reader ablations, same-history controls, and cross-person readouts support the combination of structured evidence, cross-group reader fitting, and learned multi-day organization.

AI 总结：发现事后重标记在多目标强化学习中导致偏好覆盖坍缩，提出her_mix方法恢复性能并大幅降低弃权偏好质量。

链接：https://arxiv.org/abs/2609.26918

机构：Mila (Quebec AI Institute)（米拉（魁北克人工智能研究所））； Université Laval（拉瓦尔大学）； CIFAR AI Chair（CIFAR人工智能教席）

作者：Baptiste Bonin, Caro Strickland, Audrey Durand

英文摘要：Hindsight relabeling which retroactively replacing a transition's goal with the outcome the agent actually achieved is an effective tool for improving sample-efficiency in Reinforcement Learning (RL). A natural extension to preference-conditioned multi-objective RL (MORL) relabels transitions with the preference direction the agent achieved rather than the one asked for. We show that this extension is frequently harmful: across four preference-conditioned off-policy algorithms spanning two critic backbones and two preference-sampling schemes on the continuous-control MO-Gymnasium suite, it degrades 19 of 36 algorithm-environment settings by as much as four standard deviations, improves only one, and leaves the rest unaffected. The harm is not a symptom of noisy relabels; denoising the target recovers almost nothing, and neither prioritized sampling nor any buffer-structural choice reproduces it. Instead, repeated relabeling collapses the critic's coverage onto whatever narrow region of the preference space the agent happened to visit. We name this failure mode \emph{Preference Coverage Collapse}, and quantify it with abandoned preference mass (APM), a value-aware statistic that tracks the harm ($\rho = -0.73$) where a purely structural coverage count does not. We then introduce \texttt{her\_mix}, a single-parameter convex combination pulling the achieved direction back towards the requested preference. At one fixed value across every algorithm and environment, it returns 16 of the 19 harmed settings to baseline, preserves and even improves the one setting in which relabeling helps, and cuts abandoned preference mass from $69\%$ to $6\%$. Protecting coverage over the preference simplex, not filtering noisy relabels, is what makes hindsight relabeling safe for MORL.

AI 总结：本文提出WTF框架，通过最优传输正则化将奖励微调转化为确定性最优控制问题，实现无模拟强化学习，在少步推理下提升奖励并减少280倍训练计算。

链接：https://arxiv.org/abs/2609.27033

作者：Abbas Mammadov, Jerry Y. Huang, Justin Lin, Partha Kaushik, Sheel Shah, Kartik Nair, Yee Whye Teh, Nicholas M. Boffi

英文摘要：Reward fine-tuning aims to update a pre-trained flow-based generative model to improve the downstream reward of its generated samples. Existing methods typically formulate this problem as sampling from a reward-tilted distribution, the solution to a KL-regularized reward-maximization problem. Here, we introduce an optimal transport regularizer built directly from the pre-trained drift. Unlike KL reward tilting, the resulting objective transports individual samples toward higher reward rather than reweighting the base distribution. We show that the resulting problem is equivalent to a deterministic optimal control problem on the flow. Given a pre-trained flow map, this equivalence yields a simulation-free reinforcement learning algorithm for fine-tuning generative flows. We call the resulting framework Wasserstein-Tilted Flow Maps (WTF), the first end-to-end fine-tuning recipe native to flow maps. The output is a fine-tuned flow map that retains strong reward-aligned performance at few-step inference budgets without post-hoc distillation. Experiments on ImageNet-256 and text-to-image show that WTF achieves higher reward with comparable or higher diversity than baselines, while requiring up to $280\times$ less training compute. More broadly, we argue that accelerated samplers such as flow maps are essential infrastructure for efficient post-training, and that the dominant KL-regularized formulation is only one of many choices worth revisiting.

AI 总结：提出CC-OPD反事实约束条件化在线蒸馏方法，通过消融教师条件中的每个约束构建令牌级塑形奖励，提升多约束指令遵循性能，在多个基准上取得最优平均结果，1.5B学生超越7B教师。

链接：https://arxiv.org/abs/2609.27421

机构：Alibaba Group（阿里巴巴集团）

作者：Yanzhao Zheng, Yuanqiang Yu, Tianze Xu, Chao Ma, Zhentao Zhang, Jihuai Zhu, Baohua Dong, Hangcheng Zhu, Ruohui Huang

英文摘要：Multi-constraint instruction following requires a model to respond to a query under many simultaneously active constraints. Even strong instruction-tuned models still routinely violate some of them. Existing approaches either augment supervision with sequence- or token-level RL rewards from external verifiers or learned graders, or use on-policy distillation (OPD) against a single full-context teacher whose probability mass becomes diluted as more constraints become simultaneously active. We propose CC-OPD (Counterfactual Constraint-Conditioned On-Policy Distillation), which inverts the standard supervision-generation direction in distillation. Rather than enriching the teacher with information beyond what the student sees, CC-OPD ablates each constraint from the teacher's conditioning in turn, and constructs the per-constraint signal from the resulting per-token probability differentials. The resulting per-token leave-one-out log-likelihood shifts are summed, clipped, and added to the vanilla OPD reward as a token-level shaping term. All shaping terms are obtained from the frozen teacher, without an external verifier during distillation, and the reward equals vanilla OPD wherever the aggregate shift is zero. Across two Qwen model pairs and seven benchmarks, CC-OPD achieves the highest average among all evaluated student-training methods. A 1.5B student trained with CC-OPD surpasses its own 7B RL-trained teacher on the MulDimIF benchmark.

AI 总结：提出QRLQ框架，将参数化量子电路与决斗双深度Q网络结合，在量子云编排中实现成本与延迟权衡，相比启发式方法降低成本延迟，且参数更少。

链接：https://arxiv.org/abs/2609.27446

机构：University of Information Technology, Vietnam National University, Ho Chi Minh City（胡志明市越南国家大学信息技术大学）； Monash University（莫纳什大学）； CSIRO（澳大利亚联邦科学与工业研究组织）

作者：An N. H. Phan, Dang Van Huynh, Muhammad Usman, Hoa T. Nguyen

英文摘要：Quantum cloud computing, delivered through the quantum-as-a-service (QaaS) model, provides access to quantum computing resources. However, applying uniform time-based pricing across fundamentally heterogeneous quantum resources significantly complicates task orchestration, particularly when addressing the tradeoff between execution costs and system performance. While heuristic methods rely on predefined scheduling rules, classical deep reinforcement learning (DRL) models may require more trainable parameters in this setting. Motivated by the potential of parameterised quantum circuits (PQCs) as compact function approximators, we propose QRLQ, a cost-delay-aware quantum cloud scheduling framework integrating PQCs with a dueling double deep Q-network (D3QN) to dynamically account for both cost and delay. Our simulation results show that QRLQ achieves lower mean cost and delay than the heuristic baselines, achieving a 5-11% lower mean cost relative to availability-based and rotation-based heuristics and reducing mean delay by 17% and 82% relative to the strongest and weakest heuristic baselines, respectively, while retaining execution fidelity within 2% of a fidelity-greedy policy. Compared with the classical DRL baseline, QRLQ achieves comparable scheduling performance while using 72% fewer trainable parameters. This work explores the feasibility of using QRL for task orchestration in quantum cloud environments and demonstrates its potential for cost-delay-aware quantum resource management.

AI 总结：针对长视界智能体任务中结果奖励稀疏的问题，提出ProCredit方法，通过重新运行验收检查将进度转化为信用，在多个规模上提升任务完成率。

链接：https://arxiv.org/abs/2609.27532

机构：University of Chinese Academy of Sciences（中国科学院大学）； Institute of Neuroscience, Chinese Academy of Sciences（中国科学院神经科学研究所）； Tongyi Lab, Alibaba Group（阿里巴巴集团通义实验室）； University of California, Los Angeles（加利福尼亚大学洛杉矶分校）； Nanyang Technological University（南洋理工大学）

作者：Ming Ma, Yi Zhu, Yiran Zhong, Feida Zhu, Chonghan Liu, Pengkun Jiao, Qichao Wang, Yanhao Jia, Tianming Yang, Steven Hoi

英文摘要：Long-horizon agentic tasks require an agent to modify an environment through a sequence of tool calls, with success determined by the final state. The standard recipe assigns a single outcome reward at the end and compares trajectories sampled for the same task. As a result, a group with no successful trajectory yields no training signal, failed attempts cannot be told apart by how close they came to completion, and turns that advance the task receive the same credit as turns that only query the environment. Prior work refines the unit of comparison from the trajectory to the step, or trains a reward model to supply intermediate signal: the former still derives its signal from final success alone, and the latter estimates it with a model. We observe that the acceptance checks that decide success can also be run on intermediate states, so progress is as verifiable as the outcome. We propose ProCredit, which turns this verified progress into credit: it reruns the acceptance checks after each turn, rewards the turn by its change in progress, and uses these rewards to assign credit both across attempts at the same task and across the turns within a trajectory. Starting from Qwen3.5 base models at three scales on AppWorld, ProCredit outperforms outcome-reward baselines and progress-based baselines in task completion rate at every scale on both test sets, exceeding the strongest outcome-reward baseline by 4.1 percentage points at 4B, and results in a second environment show the same direction of improvement. Ablations show that adding the final progress to the trajectory score alone does not improve performance: the gain comes from crediting progress to the turn where it occurs.

AI 总结：针对具身强化学习中资源利用不均和同步停滞问题，提出EBRL异步训练系统，通过异步流水线调度和细粒度资源管理，实现1.30-3.47倍吞吐量和2.5倍收敛速度提升。

链接：https://arxiv.org/abs/2609.27547

机构：Nanjing University（南京大学）； Institute for AI Industry Research (AIR), Tsinghua University（清华大学智能产业研究院（AIR））； Nanjing University of Posts and Telecommunications（南京邮电大学）； University of Science and Technology of China（中国科学技术大学）

作者：Liang Mi, Weijun Wang, Bowen Gao, Tianze Yu, Zixu Hao, Han Xiao, Xin Ding, Mingzhe Huang, Xin He, Lu Shi, Hao Wu, Haipeng Dai, Guihai Chen, Yunxin Liu, Ting Cao

英文摘要：Embodied reinforcement learning (RL) improves model capabilities with a pipeline of environment simulation, action generation, and model updates. These stages show heterogeneous CPU and GPU demands, making efficient resource utilization difficult. Recent systems overlap rollout (simulation and generation) with training for efficiency, but exclusive GPU allocation and synchronized barrier in rollout still leave substantial hardware resource waste. In this paper, we present EBRL, an asynchronous embodied RL training system with two core techniques. The asynchronous pipelined scheduler overlaps rollout and training, pipelines simulation and generation across environment groups, and carries out each environment independently, eliminating synchronization stalls. The fine-grained resource manager pools CPU cores and GPU streaming multiprocessors, and uses stage profiles and runtime feedback to adjust resource quotas and batch sizes to meet the shifting demands among stages. We implement EBRL on RLinf and evaluate it with four embodied policies and four simulation benchmarks across heterogeneous GPU testbeds. Experiments show that EBRL achieves 1.30-3.47 times the end-to-end rollout throughput and 2.5 times of training convergency compared to the SOTA embodied RL systems.

AI 总结：提出DCRL框架，通过策略-奖励流形对齐解决RL中奖励系统不稳定与奖励黑客问题，利用逻辑提示演化和联合更新机制，显著提升LLM推理性能，小模型超越大模型。

链接： https://arxiv.org/abs/2609.27572

机构：The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））； Tencent HY（腾讯HY）

作者：Henan Sun, Zehua Li, Haitao Hu, Qifan Zhang, Jianfeng Zhang, Nuo Chen, Jia Li

英文摘要：Reinforcement learning (RL) has emerged as a key paradigm for improving the reasoning capabilities of large language models (LLMs). However, existing reward systems, such as rule-based and reward-model-based, often exhibit issues such as unstable optimization and reward hacking. In this work, we revisit the general reasoning of LLMs from a geometric perspective, conceptualizing it as a coupled manifold composed of three interdependent sub-manifolds: logical deduction, evaluation, and representation. Based on this perspective, response generation in RL can be interpreted as a decoupling process from the evaluation manifold, while reward estimation corresponds to a decoupling process from the logical deduction manifold. The limitations of rule-based and reward-model RL systems can be geometrically interpreted as the mismatch of policy-reward manifolds during RL process. To address the aforementioned misalignment, we propose Decoupling and Coupling Reinforcement Learning (DCRL) framework, which incorporates two key components: (1) a syllogistic logic-based prompt evolution mechanism that dynamically refines reward rubrics to enhance the expressiveness of the reward manifold; and (2) a policy-reward re-coupling mechanism that jointly updates the reward and policy models, ensuring consistent evaluation and mitigating manifold mismatch during training. Theoretical analysis and extensive experiments across multiple reasoning domains demonstrate that DCRL consistently outperforms both rule-based and reward-model baselines. Notably, a Qwen3-4B model trained under DCRL surpasses a Qwen3-32B baseline and approaches the performance of a Qwen3-235B model, highlighting superior effectiveness and generalization in RL.

AI 总结：针对高维线性老虎机计算成本高的问题，提出基于聚类感知草图化的CS-LB算法，通过保留聚类协方差信息实现亚线性遗憾，并将每轮计算降至O(l^2d)。

链接：https://arxiv.org/abs/2609.27594

机构：University of Science and Technology of China（中国科学技术大学）

作者：Hantao Yang, Hong Xie, Defu Lian

英文摘要：We study the problem of computational efficiency for linear bandits in high-dimensional settings with a finite arm set. In linear bandits, the increase in the dimension $d$ of the feature vectors leads to growing computational costs of $O(d^2)$ at each round of update. Traditional sketching-based methods such as SOFUL reduce computation via fixed-size matrix sketching, yet run the risk of incurring vacuous linear regret when the spectral tail of the data is heavy and the sketch size is inadequately selected. To guarantee regret convergence and effectively reduce computational costs, we introduce a clustering mechanism and propose the Cluster Sketch Linear Bandit (CS-LB) algorithm. Our method preserves the full covariance information in each cluster to guarantee robust sublinear regret without spectral-tail vulnerabilities, performs cluster switching by assigning a sentinel for each cluster, and reduces per-round update computation to $O(l^2d)$ via a tunable sketch size $l

AI 总结：针对鲁棒对抗强化学习中的优化不稳定和价值估计退化问题，提出RACER框架，通过状态相关对抗目标和评论家一致性正则化，提升性能、鲁棒性与训练稳定性。

链接：https://arxiv.org/abs/2609.27667

机构：Shenzhen International Graduate School, Tsinghua University（清华大学深圳国际研究生院）

作者：Jiaxi Wu, Tiantian Zhang, Yuxing Wang, Yongzhe Chang, Xueqian Wang

英文摘要：Reinforcement learning (RL) achieves strong performance in sequential decision-making but remains brittle under dynamic uncertainty and distributional shifts. Robust Adversarial Reinforcement Learning (RARL) improves robustness via worst-case perturbations, but existing approaches frequently suffer from unstable optimization and degraded value estimation. In particular, overly aggressive adversaries can drive the agent toward uninformative failure states, while adversarial perturbations amplify disagreement between double critics and introduce biased value targets. We propose a unified framework, RACER (Risk-sensitive robust Adversarial critic ConsistEncy-regularized Reinforcement learning), that revisits adversarial RL from a risk-sensitive perspective. First, we introduce a state-dependent adversarial objective that adaptively regulates perturbation strength, suppressing harmful disturbances while preserving informative exploration. Second, we propose critic consistency regularization to reduce disagreement between Q-value estimators and stabilize learning. Comprehensive experiments on challenging continuous control benchmarks demonstrate that RACER consistently improves performance, robustness, and training stability over strong robust RL baselines.

AI 总结：本文提出极限核 Q($λ$)（LKQL）离策略价值估计器，结合 $n$ 步截断与极限核长视界近似，在保持计算复杂度的同时加速策略评估，并在 MuJoCo 基准上验证其优于 $n$ 步基线，尤其擅长长视界任务。

链接：https://arxiv.org/abs/2609.27741

机构：Delft University of Technology（代尔夫特理工大学）； Alpha Brain Technologies； University of Toronto（多伦多大学）

作者：Tolga Ok, Arman Sharifi Kolarijani, Peyman Mohajerin Esfahani, Mohamad Amin Sharifi Kolarijani

英文摘要： In value-based reinforcement learning, improving the accuracy of policy evaluation has been shown to improve downstream policy optimization performance. The widely adopted family of approximations relying on $n$-step truncation yields computationally efficient value estimators but is inherently limited to a short evaluation horizon. In contrast, methods that exploit the global structure of the transition dynamics can accelerate policy evaluation, but their memory and computational requirements often limit scalability to large or continuous state spaces. To reconcile these limitations, we introduce Limiting-Kernel Q($\lambda$) (LKQL), an off-policy value estimator that combines $n$-step truncation with a long-horizon approximation based on the limiting kernel (LK). LKQL has the same order of complexity as $n$-step estimators and integrates directly into both on- and off-policy actor-critic algorithms. We prove that, under aperiodicity and in the near-on-policy regime, the operator underlying LKQL improves the policy evaluation convergence rate over its truncated counterpart for sufficiently large $n$, and that LKQL itself converges almost surely to the optimal values in finite Markov decision processes (MDPs) under a fixed behavior policy. On the MuJoCo continuous-control benchmark, we show that LKQL improves over $n$-step baselines in most settings, particularly on long-horizon tasks.

AI 总结：本文提出验证器认证规则迁移（VCRT），通过反菱形监督保留真实先决条件，在多个推理环境中将宏通过率提升13.06个百分点，但未证实精确轨道聚合的普遍优势。

链接：https://arxiv.org/abs/2609.27833

作者：Bang Xie, Hao Liu, Zhiyuan Peng, Xin Yin, Chenhao Ying, Yuan Luo, Senjian Zhang, Wei Chen

英文摘要：Many computations admit several valid execution orders because independent subgoals or disjoint state updates can commute. Reinforcement learning with verifiable rewards usually treats each successful trace as a separate token sequence, so serialization choices can be mistaken for logical dependencies. We introduce Verifier-Certified Rule Transport (VCRT), which replays adjacent operation pairs with native verifiers. Pairs whose two orders are accepted and reach the same canonical state provide commutation certificates; rejected or state-changing reversals provide anti-diamonds. VCRT uses anti-diamonds to preserve genuine prerequisites and assigns policy credit to the total probability mass of each certified orbit. It also constrains post-swap consistency, source retention, and policy drift. We evaluate leave-one-environment-out transfer across ProofWriter, CLRS, and Lean through a shared anonymized relation-graph interface. All training and checkpoint decisions are frozen before held-out evaluation, which uses one greedy trajectory per item without search or verifier feedback. VCRT obtains a 77.60% macro pass rate versus 64.53% for the strongest matched baseline, a paired gain of 13.06 points (95% bootstrap CI [12.58, 13.54]). Lean accounts for most of this gain at 33.49 points, while ProofWriter and CLRS improve by 2.85 points on average. Mechanism tests consistently favor anti-diamond supervision, whereas No-Orbit is statistically indistinguishable from full VCRT. The evidence does not establish a general benefit from exact orbit aggregation.

AI 总结：PCQC利用特权患者信息构建反事实问题答案，通过诊断评分器分配问题信用，在四个医疗基准上以更少询问轮次显著提升诊断准确率。

链接：https://arxiv.org/abs/2609.27987

机构：Peking University（北京大学）

作者：Chenxuan Li, Jiayi Wan, Xinrong Chen, Zhongyu Zhao, Xuecheng Shang, Peixing Wan

英文摘要：Large language models (LLMs) have made substantial progress on medical question-answering, yet effective medical dialogue also requires learning to ask questions that uncover relevant patient information. To train such dialogue policies, a common pipeline combines supervised fine-tuning with reinforcement learning (RL) based on final diagnostic correctness. However, this outcome-based supervision does not directly distinguish the contributions of individual questions and provides no question-level feedback for unexecuted alternatives. To address this gap, we introduce PCQC (Privileged Counterfactual Question Credit), which uses privileged patient information during training to learn from questions never asked. During training, PCQC makes alternative questions directly comparable at the same dialogue state by using privileged patient facts to construct their answers. A frozen diagnostic scorer evaluates the diagnostic utility of each resulting question-answer pair by how strongly it supports the correct diagnosis. PCQC turns these comparisons into relative question credit that teaches the policy which questions to favor, directly supervising both executed and unexecuted questions alongside outcome-based RL without requiring complete rollouts for the unexecuted alternatives. Extensive experiments across four medical benchmarks demonstrate that PCQC achieves 63.10% mean diagnostic accuracy, outperforming GRPO and ATPO by 4.38 and 4.21 percentage points, respectively. These gains are achieved with 33.1% fewer inquiry turns than GRPO.

AI 总结：本文研究基于图神经网络强化学习的课程学习在作业车间调度中的应用，通过逐步从小实例训练到大实例，显著降低训练时间并提升泛化性能，在30×30目标规模下平均最优性差距降低约8个百分点。

链接：https://arxiv.org/abs/2609.28085

机构：Rosenheim University of Applied Sciences（罗森海姆应用科学大学）

作者：Jayakrishnan K. Vasudevan (Rosenheim University of Applied Sciences), Jonathan Hoss (Rosenheim University of Applied Sciences), Noah Klarmann (Rosenheim University of Applied Sciences)

英文摘要：The job shop scheduling problem is a challenging combinatorial optimization problem, and recent reinforcement learning approaches using graph neural networks have shown promise for learning scheduling policies directly from problem instances. However, training on large instances remains computationally expensive, and generalization across instance sizes remains challenging. This paper studies curriculum learning for graph neural network-based reinforcement learning in the job shop scheduling problem by comparing it with single-size training across three target sizes: 20 x 20, 25 x 25, and 30 x 30. In the curriculum setting, the policy is first trained on smaller instances and then progressively adapted to larger target sizes, allowing scheduling behavior learned in earlier stages to support learning on larger instances. Models are evaluated on unseen instances from 8 x 8 to 30 x 30 using the optimality gap, considering both generalization across all evaluation sizes and specialization on the target size. Results show that curriculum learning consistently reduces wall-clock training time, with larger benefits as the target size increases. The strongest advantage is observed at 30 x 30, where curriculum learning reduces the mean optimality gap across all evaluation sizes by approximately 8.1 percentage points, reduces the target-size mean optimality gap by approximately 8.6 percentage points, and saves approximately 50 hours of training time.

AI 总结：本研究探讨同策略蒸馏作为强化学习准备阶段的价值，发现其能提升最终性能，且最佳蒸馏目标取决于轨迹来源和后续训练。

链接：https://arxiv.org/abs/2609.28145

机构：Fudan University（复旦大学）； SII； Peking University（北京大学）； The Hong Kong University of Science and Technology（香港科技大学）； The Chinese University of Hong Kong（香港中文大学）

作者：Shuai Dong, Yongfu Zhu, Yuqi Xu, Weichu Xie, Liuwenpu, Ziyue Wang, Kaiwen Tuo, Congcong Wang, Siyuan Wang, Wenqi Shao, Shuai Yang, Ji Zhao, Caoyuan Ma, Wenzheng Chang, Taiqiang Wu, Xinlei Yu, Hongrui Wu, Xiaoxuan He, Fangke Chen, Dianyi Wang, Kanghui Tian, Sirry Chen, Xingyu Liu, Xiangnan Wu, Jiawei Guo, Haowen Hou, LingHan Chen, Zhongyu Wei, Jiaqi Wang

英文摘要：Reinforcement learning (RL) improves reasoning, but its performance depends on the policy from which training begins. We study on-policy distillation (OPD) as a preparation stage for RL and ask whether its benefits extend beyond improvements in the distilled model's initial accuracy. Under shared RL settings, students initialized with OPD reach higher final performance than those trained with direct RL or supervised fine-tuning followed by RL. This advantage can emerge even when OPD produces little immediate improvement in accuracy. Pre-RL Pass@k does not fully explain the benefit: similar or even higher values do not necessarily lead to better performance after RL. Behavioral analyses point to alignment with the teacher's distribution beyond top-1 agreement as a possible explanation. Such alignment may favor higher-quality reasoning paths while retaining alternatives that RL can further refine using outcome feedback. We further examine how trajectory sources and divergence objectives affect the value of distillation for subsequent RL. Standard reverse-KL OPD performs better before RL, but forward-KL OPD overtakes it afterward; with teacher-generated distillation trajectories, reverse KL remains ahead at both stages. These findings suggest that the preferred distillation objective depends on both the trajectory source and the training that follows. Our results support evaluating OPD as preparation for RL and selecting distillation choices by the performance achieved after subsequent training.

AI 总结：提出UECR-GRPO，通过熵校准信用分配统一在线蒸馏与GRPO，在响应和词元级别整合验证器与教师信号，在五个数学推理基准上超越最强基线。

链接：https://arxiv.org/abs/2609.28385

机构：Shanghai Jiao Tong University（上海交通大学）； Zhejiang University（浙江大学）

作者：Jie Zhang, Jingxiao Yang, Zhehao Huang, Yuhang Liu, Xiaolin Huang

英文摘要：Reinforcement learning with verifiable rewards (RLVR) supervises mathematical reasoning through final-answer correctness, but provides little guidance on individual tokens. On-policy distillation (OPD) supplies dense feedback on student-generated responses, yet teacher preference need not reflect correctness. Recent hybrids combine OPD and verifier-derived advantages or reweight task credit using teacher ratios. However, teacher guidance enters after verifier-based group normalization, and token reweighting need not preserve the total task credit assigned to each response. We introduce Unified Entropy-Calibrated Credit Redistribution for GRPO (UECR-GRPO), which integrates verifier and teacher signals within a single GRPO-style update at both the response and token levels. \emph{Path-Utility Unification} (PUU) combines verifier reward and a teacher-to-anchor path log-ratio in a single KL-regularized objective. Its on-policy implementation uses a length-normalized teacher score and combines both rewards before group normalization and PPO clipping, allowing teacher evidence to influence the response ranking. \emph{Entropy-Calibrated Redistribution} (ECR) then uses the signed teacher--old-policy token gap to redistribute the verifier-derived component. Full-vocabulary teacher entropy attenuates uncertain guidance, while a response-wise zero-sum projection preserves the total task credit and its token-wise sign before clipping. Across five mathematical reasoning benchmarks, UECR-GRPO achieves average \(\mathrm{Avg@12}\) accuracies of 17.21\% and 65.09\% with Qwen3-1.7B and Qwen3-4B students, respectively, exceeding the strongest baseline at each scale by 0.89 and 0.56 percentage points.

AI 总结：本文提出两种分布式递归高斯过程算法ADMM-RGP和PDMM-RGP，用于多智能体系统中的多输出回归，通过参数选择加速收敛以减少通信，并在真实风场数据上验证了其高效性和准确性。

链接：https://arxiv.org/abs/2609.26979

机构：Delft University of Technology（代尔夫特理工大学）

作者：Josephine King, Ali Emre Balci, Raj Thilak Rajan

英文摘要：Gaussian processes (GPs) provide a flexible framework for learning unknown functions from noisy measurements while quantifying predictive uncertainty, making them well suited for estimation in multi-agent systems. However, when measurements are collected by multiple agents, maintaining a unified GP model without centralized processing requires efficient distributed algorithms that can operate using local measurements and communication with neighboring agents. In this work, we develop two distributed recursive GP (RGP) algorithms for multi-output GP regression: ADMM-RGP and PDMM-RGP. We analyze the stability and convergence of both algorithms and develop parameter selection strategies to accelerate convergence, thus reducing the communication burden. The proposed methods are validated on a real-world multi-output wind dataset, and their convergence behavior is examined across communication graphs with varying connectivity. Numerical experiments demonstrate that ADMM-RGP and PDMM-RGP can significantly reduce communication relative to the state of the art, while maintaining comparable estimation accuracy and network-wide consensus.

AI 总结：本文提出LLMAE，通过暴露中间固定长度潜瓶颈，将预训练解码器语言模型重用作连续文本自编码器，实现高保真重建并支持下游潜文本扩散模型用于图像描述。

链接：https://arxiv.org/abs/2609.27248

机构：University of California, Irvine（加州大学尔湾分校）

作者：Arkanath Pathak, Unnat Jain, Alexander C. Berg

英文摘要： Next-token prediction has enabled highly fluent autoregressive language models, but it represents global structure only indirectly through sequential factorization. In contrast, high-fidelity autoencoders have become a standard primitive in image generation, enabling generative models to operate over continuous latent spaces; text lacks a comparably faithful continuous representation. We propose LLMAE, a method for repurposing a pretrained decoder-only language model as a continuous text autoencoder by exposing an intermediate fixed-length latent bottleneck within its internal activations. Instantiated with a parameter-efficient 270M Gemma 3 model, LLMAE uses structured attention masks, LoRA adaptation, and KL regularization to learn an autoencoding interface that leverages the generative prior of the original LLM. We train LLMAE to reconstruct text sequences up to 1024 tokens, significantly improving on this task to achieve near-perfect reconstruction. Furthermore, we demonstrate the downstream utility of this representation by training a latent text diffusion model for detailed image captioning using the learned LLMAE autoencoder. By mapping text into a fixed-length continuous latent space, our approach provides an effective substrate for downstream adaptation while benefiting from the fluency of the original LLM.

AI 总结：本文提出一种基于变分自回归网络的离散扩散模型，通过显式马尔可夫跳跃算子实现归一化分布，应用于高维伊辛模型，准确计算热力学量，并与蒙特卡洛采样结合提升低温采样效率。

链接：https://arxiv.org/abs/2609.27306

机构：Institute of Fundamental and Frontier Sciences, University of Electronic Science and Technology of China（电子科技大学基础与前沿研究院）； School of Physics, University of Electronic Science and Technology of China（电子科技大学物理学院）； Key Laboratory of Quantum Physics and Photonic Quantum Information, Ministry of Education, University of Electronic Science and Technology of China（电子科技大学教育部量子物理与光子量子信息重点实验室）； Non-classical Information Science Basic Discipline Research Center of Sichuan Province, University of Electronic Science and Technology of China（电子科技大学四川省非经典信息科学基础学科研究中心）

作者：Kewen Pan, Ying Tang

英文摘要：Conventional score-based diffusion models learn scores without representing normalized densities, whereas tractable normalized models support both sampling and direct likelihood evaluation. A recent tensor-network approach provides such a representation but is largely restricted to low-dimensional lattices. Here we introduce a discrete diffusion model that parameterizes normalized probability distributions using variational autoregressive networks. Explicit Markov jump operators govern the forward noising and reverse denoising dynamics, extending discrete diffusion models with normalized distributions to spin systems on higher-dimensional lattices. We apply this framework to the two- and three-dimensional Ising models across ordered, critical, and disordered regimes, accurately computing thermodynamic quantities including free energy, energy, and magnetization. We further integrate the framework with Monte Carlo sampling, using adaptive diffusion steps to maintain high acceptance rates even at low temperatures while enhancing sample diversity. These results establish a neural-network framework for the discrete diffusion model with normalized probability distributions.

AI 总结：针对托卡马克边界等离子体模拟，提出将曲线网格展开为图像张量并采用条件流匹配模型，实现概率性代理，在敏感工作点捕获多种结果而非平均。

链接：https://arxiv.org/abs/2609.28116

机构：Eindhoven University of Technology（埃因霍温理工大学）； DIFFER（荷兰等离子体物理研究所）

作者：Gabriele Gianuzzo, Stefan Dasbach, Fleur Hendriks, Sven Wiesen, Vlado Menkovski

英文摘要：Fast surrogates for tokamak boundary-plasma simulation are typically deterministic regressors mapping a global operating point to a flattened vector of cell values. Near the divertor detachment transition the steady state is not reliably single-valued. A point estimate must average over qualitatively different plasma states, and it arrives with no statement of confidence. Moreover, the flattened vector representation discards the geometric structure of the SOLPS-ITER mesh. This work addresses both problems. We unroll the curvilinear mesh into three fixed-size image tensors whose layout preserves cell adjacency and inverts exactly, letting a convolutional network act on the geometry without loss of information. A conditional flow matching model, well suited to highly sensitive systems, is then trained on this representation. The result is an efficient, scalable surrogate that captures multiple plausible outcomes even at sensitive operating points. Along a gas-puff scan, the predictive distribution splits into a hot and a cold mode across an early regime transition. A further check on synthetic data with an injected bifurcation of known size confirms the model recovers both branches rather than their average.

AI 总结：本文提出Clifford-VAE，一种将数据投影到任意维度克利福德环面的变分自编码器，用于学习全息约简表示，在半监督分类中与高斯和超球面VAE竞争，并在VSA基准测试中表现更优，为感知数据接地到符号推理提供了新方法。

链接：https://arxiv.org/abs/2609.28409

机构：Institute of Neuroinformatics（神经信息学研究所）； University of Zurich & ETH Zurich（苏黎世大学与苏黎世联邦理工学院）； National Research Council Canada（加拿大国家研究委员会）； University of Waterloo（滑铁卢大学）

作者：Mohamed Malek Abid, P. Michael Furlong

英文摘要：Vector Symbolic Algebras project data structures into a hyperdimensional vector space through the application of their vector algebras to randomly generated atomic vector symbols and fractional power encodings of real-valued data. Embedding unstructured data remains an open question. We present \textit{Clifford-VAE}, a variational autoencoder that learns to project data onto a Clifford torus in arbitrary dimensions. Experiments using the MNIST, FashionMNIST, and CIFAR-10 datasets demonstrate that Clifford-VAE produces representations that are competitive with those produced by Gaussian and Hyperspherical VAEs for semi-supervised classification tasks while outperforming Gaussian and Hyperspherical counterparts in the VSA benchmark tests of self-binding and unbinding, role-filler recovery, and bundle capacity. Clifford-VAE provides a principled technique for grounding perceptual data into a symbolic reasoning framework, providing a new approach to a long-standing problem in the VSA literature.

AI 总结：提出上下文连续偏好学习（CCPL），利用高斯过程在邻近操作条件间共享偏好数据，在有限反馈下提升外骨骼个性化效率，模拟和人体数据验证了其有效性。

链接：https://arxiv.org/abs/2609.28427

机构：Korea University（高丽大学）； WIRobotics

作者：Sunin Baek, Sungwoo Park, Daekyum Kim

英文摘要：Personalizing exoskeleton assistance across operating conditions is constrained by the time and physical effort required to collect user feedback. We examined whether a user's preference landscape varies smoothly across operating conditions and when this continuity supports learning from limited feedback. We propose Context-Continuous Preference Learning (CCPL), a Gaussian-process preference model that shares observations across nearby contexts while retaining context-specific utility estimates. We evaluated CCPL through simulations and retrospective analyses of ankle and elbow exoskeleton preference data from nine healthy adults. In simulations, CCPL improved reconstruction and preference-based Bayesian optimization relative to independent learning when preferences varied smoothly, but showed negative transfer when continuity was weak. In both human studies, full-data reference landscapes estimated separately for each participant and context tended to be more similar between nearby operating conditions. With five exposures per context, CCPL increased mean reconstruction correlation with these references from 0.644 to 0.720 for ankle assistance and from 0.476 to 0.526 for elbow assistance relative to independent learning. The five-exposure budget was approximately 37% lower for ankle and 17% lower for elbow than the estimated independent-learning budgets needed to match these correlations. CCPL also improved held-out response prediction relative to independent learning, while benefits over pooled learning varied. These findings support context continuity as a basis for sharing preference observations under limited feedback, although benefits for online personalization in humans remain to be established.

AI 总结：针对稀缺数据下的图学习问题，提出SCoGL框架，通过拉普拉斯谱先验促进全局连通性，结合投影梯度下降优化，实验证明能提升图恢复和下游任务性能。

链接：https://arxiv.org/abs/2609.27278

机构：Tsinghua University（清华大学）； York University（约克大学）

作者：Mingxiao Liu (1), Bahar Oveisgharan (2), Bingyan Zou (1), Gene Cheung (2), H. Vicky Zhao (1), Feifei Gao (1) ((1) Tsinghua University, China, (2) York University, Canada)

英文摘要：Learning a sparse graph from scarce data is practically important but challenging. Motivated by the desirable combination of local sparsity and strong global connectivity exhibited by expander-like graphs, we propose spectral connectivity-regularized graph learning (SCoGL), a framework that incorporates a family of Laplacian spectral priors to explicitly promote global connectivity. Specifically, SCoGL augments a combinatorial-Laplacian-constrained graphical lasso (GLASSO) objective over a target adjacency matrix $\mathbf{W}$ with a general connectivity prior computed from Laplacian eigenvalues. We derive gradients for several representative connectivity priors and develop a projected gradient descent (PGD) algorithm with Armijo backtracking to efficiently optimize $\mathbf{W}$. Experiments show that the proposed SCoGL variants improve graph recovery and enhance downstream tasks such as graph signal denoising when signal observations are scarce.

AI 总结：本文揭示线性擦除在密集叠加下失效，网络通过镜像或阴影解非线性抑制特征，留下可恢复痕迹，解释了LLM遗忘中抑制而非删除的失败机制。

链接：https://arxiv.org/abs/2609.27593

机构：Techno India University（泰克诺印度大学）； DRDO Young Scientist Laboratory - CT（DRDO青年科学家实验室-CT）

作者：Akash Samanta, Manish Pratap Singh, Debasis Chaudhuri

英文摘要：Concept erasure methods that operate via linear projection assume that features occupy separable subspaces. We show this assumption fails under dense superposition: when two features are forced into an antipodal pair sharing a single subspace, state-of-the-art linear erasure destroys both, not just the target. Networks trained with gradient descent instead solve this problem non-linearly, but not uniformly: they converge to one of two distinct circuit-level solutions depending on initialization, which we call mirror and shadow solutions. We map this bifurcation as a function of feature entanglement, show it reflects a stable attractor structure rather than an artifact of our setup, and use targeted causal interventions to demonstrate that both solutions leave a substantial, measurable trace of the erased feature's representation intact, recoverable through a single scalar patch rather than requiring any further training. This mirrors a failure mode recently observed empirically in LLM unlearning, where suppression rather than deletion allows forgotten knowledge to resurface; our results offer a mechanistic, causally-validated account of why that failure mode occurs.

AI 总结：Pheno-GS提出测地线Sinkhorn方法，结合图正则化、非平衡OT与批量热扩散，实现大规模患者单细胞数据的快速准确距离计算。

链接：https://arxiv.org/abs/2609.27633

机构：University College London（伦敦大学学院）； Yale University（耶鲁大学）

作者：Alistair Wilkinson, Christopher J. Tape, Smita Krishnaswamy

英文摘要：High-throughput single-cell data is now collected across large patient cohorts. Understanding patient-level heterogeneity from cellular-level data motivates phenoscaping: embedding each single-cell distribution as a "datapoint," with distances given by optimal transport (OT). Computing geometry-aware OT at this scale, between all pairs of patient datasets, remains an open challenge, since existing methods either rely on Euclidean ground metrics that distort manifold structure or fail under sparse, unevenly sampled, or large-scale data. We present \textbf{Pheno-GS} (Phenoscape-scale Geodesic Sinkhorn), which computes accurate, scalable geodesic transport distances under noisy, unbalanced, large-scale settings via three components: ($1$) graph connectivity regularization for well-defined geodesics on sparse/disconnected manifolds; ($2$) an unbalanced OT formulation via KL marginal penalties; and ($3$) a batched matrix algorithm computing all pairwise distances in one heat diffusion (over $200 \times$ faster than Geodesic Sinkhorn for $500$ distributions). We validate Pheno-GS on synthetic benchmarks and a CyTOF perturbation dataset.

AI 总结：本研究通过教师-学生线性RNN模型，推导出序列预训练中的显式缩放定律，揭示序列长度与序列数量在双尺度机制下不可互换，并给出理论证明。

链接：https://arxiv.org/abs/2609.27964

机构：The University of Sydney（悉尼大学）； Together AI； The Pennsylvania State University（宾夕法尼亚州立大学）

作者：Ziyan Chen, Zhongzhu Zhou, Peilin Liu, Ding-Xuan Zhou

英文摘要：Empirical scaling laws for autoregressive language models relate prediction loss to model size, data size, and optimization compute, but their theoretical origin is still poorly understood in sequential pretraining settings. We study this question in a tractable teacher--student model where a stable latent linear RNN generates trajectories and a sketched linear recurrent student is trained by safeguarded full-batch WSD gradient descent on next-token prediction. The sketch dimension $M$ plays the role of model size, while $N$ independent trajectories of length $P$ provide the training tokens. We allow the innovation and initialization covariances to have different power-law exponents $\alpha$ and $\theta$. The induced design spectrum produces explicit approximation, optimization, and statistical scaling laws separated by spectral crossovers. When $\theta\ge\alpha$, the original one-scale rates $M^{1-\beta_\alpha}$, $R^{(1-\beta_\alpha)/\alpha}$, and $(NP)^{-1}\min\{M,R^{1/\alpha}\}$ are recovered. When $\alpha-2r\le\theta

AI 总结：提出SCFF框架，通过支持排序特征路由和有界叶节点编码，将表格基础模型的二次方特征交互降为线性计算，在18数据集上以更低内存提升准确率，相对误差降低最高26.1%。

链接：https://arxiv.org/abs/2609.28208

机构：Ant Group（蚂蚁集团）

作者：Tian Zhou, Beverly Jin, Xue Wang, Linxiao Yang, Wenwei Wang, Bingqing Peng, Mengni Ye, Jinjie Gu, Liang Sun

英文摘要：Tabular foundation models face a feature-side scaling dilemma: full-width pairwise mixing grows quadratically with the number of columns, whereas feature selection saves memory by discarding evidence. We introduce Support-Compiled Feature Folding (SCFF), a training-free inference framework that resolves this dilemma without changing the frozen backbone. SCFF routes support-ranked features through bounded leaves of the native feature encoder, support-checks the residual evidence, and merges the encoded messages before a single contextual prediction. It thereby converts quadratic feature-interaction work into linear-in-width work with a bounded local working set, without ensembling predictions or training new parameters. On the exhaustive 18-dataset wide-table slice of fixed AMLB-29, TabZilla, and TabArena snapshots, SCFF improves dataset-macro accuracy and NLL on all six evaluated backbones. All four matched-width comparisons retain favorable 95 percent dataset-bootstrap intervals on locked folds, with relative error reductions up to 26.1 percent. Median paired GPU-memory savings are 2.09x to 2.36x, and the ratio of separately observed maximum peaks reaches 34.3x. Under a measured peak-memory ceiling, SCFF uses the saved budget to preserve more support-selected evidence, improving accuracy by 4.06 and 3.72 points over the widest feasible single leaf on predeclared wide-Core strata of TabICLv2 and TabPFN-3.

AI 总结：针对大规模在线线性规划，提出资源自适应随机梯度下降算法，以O(m)复杂度实现无需重新求解的库存定价，达到O(log T)遗憾，兼顾效率与分配质量。

链接：https://arxiv.org/abs/2609.28263

机构：Fudan University（复旦大学）

作者：Jiameng Lyu

英文摘要：The growth of large language model (LLM) inference and search services increases the scale of online linear programming problems, motivating computationally efficient algorithms. We develop resource-adaptive stochastic gradient descent (RASGD) for stochastic online linear programming. The algorithm uses one request and current inventory to update resource prices, requiring O(m) operations for m resources and memory per arrival and no LP or sample-average optimization. The central idea is to express the current-resource pricing logic of re-solving through a first-order SGD update: each arrival refreshes the remaining-inventory allowance in the dual objective, while the stepsize decreases for early learning and increases later to match the speed of inventory adjustment. Under standard non-degeneracy conditions, our algorithm is feasible on every sample path and achieves O(\log T) expected regret against the realized fractional hindsight optimum, which matches the lower bound, even for policies that know the distribution and have unrestricted computation. The analysis converts curvature around the fixed reference price into inventory stability without tracking optimal prices at changing resource levels. Numerical experiments show that RASGD achieves regret competitive with per-arrival LP re-solving and improves upon the tested first-order baselines, while retaining the computational efficiency of first-order methods. These results establish RASGD as a computationally efficient approach to achieving high allocation quality in large-scale OLP.

AI 总结：针对联邦学习中的隐蔽后门攻击，本文提出FedMAST，利用结构、频谱和历史证据的多轴追踪检测后门，在多种攻击下显著降低ASR并保持高主任务准确率。

链接：https://arxiv.org/abs/2609.27760

机构：Kennesaw State University（肯尼索州立大学）

作者：Srinivasan Subramanian, Kazi Aminul Islam, Md. Abdullah Al Hafiz Khan

英文摘要： Federated learning enables distributed training of a shared model without requiring clients to share their raw data. However, its reliance on the integrity of the client-submitted updates exposes the global model to stealthy backdoor poisoning. Although existing defenses often inspect isolated evidence sources, stealth-constrained attacks can adapt to these signals. In this paper, we show that such attacks can suppress isolated anomaly signals, but their poisoned updates still leave residual structural traces. We propose FedMAST, a Federated Multi-Axis Structural Tracing defense for backdoor detection in federated learning. FedMAST scores client updates using complementary structural, spectral, and historical evidence and then applies tiered filtering and round-level containment to limit adversarial influence. To capture traces that isolated signals may miss, FedMAST uses squeeze-pair coherence scoring to expose coupled feature distortions and signed spectral-drift tracking to reveal persistent directional changes over time. Across six federated backdoor attacks, namely Constrain-and-Scale, Neurotoxin, BC-Layers, LGA, DBA, and 3DFed, FedMAST achieves lower ASR than baseline defenses in all nine evaluated attack--defense comparisons. Across the complete 200-round runs, it attains an average ASR of 1.51% while maintaining 94.84% average main-task accuracy. Under the method-aware CovertLayers attack, FedAvg, MultiKrum, AlignIns, and FLAME yield full-run ASRs of 100.00%, 99.67%, 99.53%, and 32.84%, respectively. FedMAST achieves the lowest ASR among all evaluated methods, reducing it to 1.53% while maintaining 92.26% main-task accuracy.

AI 总结：针对联邦学习中特征级缺失问题，提出Fed-ReMasker方法，将掩码自编码器适配至联邦框架，利用跨中心知识填补未观测特征，在多种基准中显著优于现有方法，接近集中式模型性能。

链接：https://arxiv.org/abs/2609.28105

机构：ARTORG Center for Biomedical Engineering Research University of Bern（伯尔尼大学生物医学工程研究中心）； Graduate School for Cellular and Biomedical Sciences University of Bern（伯尔尼大学细胞与生物医学科学研究生院）

作者：Ioannis Papathanail, Rooholla Poursoleymani, Lubnaa Abdur Rahman, Stavroula Georgia Mougiakakou

英文摘要：Multi-center clinical studies and biomedical research collaborations increasingly seek to utilize data across centers to build models that generalize beyond any single center. This creates two distinct challenges: data protection regulations may restrict the sharing of raw patient data across institutions, while centers may collect only partially overlapping sets of features under different protocols. Federated learning enables collaborative model training without centralizing raw data. However, existing federated imputation methods rarely evaluate feature-level missingness, in which entire features are unobserved at some centers. To address this setting, we adapt the ReMasker masked autoencoder to federated learning (Fed-ReMasker), enabling centers to impute features never observed locally by leveraging knowledge learned across collaborating centers. We evaluate Fed-ReMasker in a benchmark spanning synthetic datasets with linear and nonlinear relationships and real-world tabular datasets, including clinical data. The benchmark varies the number of centers, the missingness ratios, and client heterogeneity. Fed-ReMasker achieves the lowest imputation error in 93.2% of value-level and 96.7% of feature-level scenarios in the homogeneous benchmark. It also remains robust to client heterogeneity using simple federated averaging, outperforming all baselines in all 36 value-level scenarios and each baseline in at least 35 of 36 feature-level scenarios, and comes within 3.0% on average of a centralized model trained on the pooled data.

AI 总结：针对校准标签含噪的表格分类器，提出LWCal及其变体Gated-LWCal，通过损失加权和保守门控降低校准误差，实验证明优于现有方法。

链接：https://arxiv.org/abs/2609.26839

机构：Brown University（布朗大学）； Georgia Institute of Technology（佐治亚理工学院）

作者：Zeming Liu, Hang Lyu, Jingtao Zhang, Yuan Xie

英文摘要：Post-hoc probability calibration is usually evaluated under an optimistic assumption: the held-out calibration labels are clean. In many AI deployment settings, however, labels come from weak annotators, historical decisions, heuristics, or distant supervision, so the same label noise that corrupts training also corrupts calibration. We study this overlooked failure mode for tabular classifiers and propose LWCal, a CPU-only post-hoc calibrator that down-weights calibration examples whose noisy labels are contradicted by the base model's held-out probability. LWCal requires no clean validation labels, no noise-rate estimate, and no retraining of the base classifier. A second variant, Gated-LWCal, adds a conservative disagreement gate that backs off toward the raw score when the calibration split appears extremely inconsistent. On nine local binary tabular tasks, six random seeds, symmetric and asymmetric label corruption, and three tree-based base learners, LWCal obtains the lowest average calibration error while Gated-LWCal obtains the best average proper-score tradeoff. In the main random-forest study over 432 noisy cells, Gated-LWCal reduces expected calibration error from 0.188 to 0.122 and negative log likelihood from 0.438 to 0.396 relative to the raw classifier. Paired bootstrap intervals for Gated-LWCal versus raw, Platt, isotonic, and beta calibration exclude zero on ECE, Brier score, and NLL. The artifact contains all scripts, result tables, figures, and the compiled paper.

AI 总结：本文提出一种尾部感知的共形椭球几何学习框架，通过两分割设计解耦尾部敏感性与覆盖保证，在CVaR约束下最小化体积，实现高效且鲁棒的多元预测集。

链接：https://arxiv.org/abs/2609.27221

机构：Nanyang Technological University（南洋理工大学）

作者：Xiang Zhang

英文摘要：This paper studies multivariate conformal prediction (CP), a distribution-free uncertainty quantification framework with finite-sample coverage guarantees. The efficiency of multivariate prediction sets hinges critically on the residual geometry encoded by the nonconformity score, while existing minimum-volume methods rely on quantile thresholds that ignore tail residual severity and implicitly bind geometry learning to coverage level. We propose a tail-aware geometry learning framework for conformal ellipsoids that decouples tail sensitivity in geometry learning from the final coverage guarantee. Using a two-split design, we learn the metric matrix via volume minimization under a CVaR constraint on an estimation split, then apply standard conformal calibration on a held-out calibration split. The resulting problem is convex and admits a bounded-reweighting interpretation that prioritizes high-residual samples. Moreover, we theoretically characterize the trade-off between ellipsoidal volume and tail severity. Experimental results demonstrate the effectiveness of the proposed method.

AI 总结：本研究通过受控实验揭示fMRI基础模型的扩展规律，发现数据和模型规模应同步扩展，且数据增加比模型规模更有效，并据此在固定计算预算下选出性能最优的模型组合。

链接：https://arxiv.org/abs/2609.27232

机构：Southern University of Science and Technology（南方科技大学）； Shenzhen University（深圳大学）； University of Warwick（华威大学）； SLAI

作者：Wenhao Ye, Xuanye Pan, Junfeng Xia, Junxiang Zhang, Mo Wang, Quanying Liu

英文摘要：Scaling laws have guided large-model development in computer vision and natural language processing, but the relationships among data, model size, and compute remain unclear for functional magnetic resonance imaging (fMRI) foundation models. Here, we conduct a controlled empirical study using pretraining data from more than 200 source datasets and over 10,000 GPU-hours of experiments. Holding the pretraining framework and downstream protocol fixed, we vary pretraining data size, model size, and training duration. Downstream performance generally improves with compute, yet models using similar compute can perform substantially differently. Additional pretraining data bring larger gains at larger model sizes, suggesting that data and model size should be scaled together. At matched compute, increasing pretraining data benefits more tasks than increasing model size, although the pattern varies across tasks. We then use in-distribution (ID) downstream performance to select the combination of pretraining data size, model size, and training duration at two fixed compute budgets. The resulting models are locked before out-of-distribution (OOD) evaluation. They achieve the highest average performance across the evaluated OOD tasks among the compared fMRI foundation models while using less pretraining compute. Overall, our results show that compute alone does not characterize fMRI scaling: performance depends on how pretraining data, model size, and training duration are combined.

AI 总结：本文系统审计了LLM裁判跨域重校准中准确率差距的预测与认证能力，发现其仅提供弱预警信号，无法作为部署证书，并提出了有限样本下界证书方法。

链接：https://arxiv.org/abs/2609.27954

机构：Kalinga Institute of Industrial Technology（卡林加工业技术学院）； Iowa State University（爱荷华州立大学）

作者：Fariya Afrin, Ibne Farabi Shihab

英文摘要：A scalar recalibration map fitted for an LLM judge on one task can fail when the task distribution changes, but the source-target accuracy gap is often treated as a proxy for that failure. We test what this gap can predict and what it can certify across thirteen judges, two generators, eight domains, and 1,176 predeclared transfers. After accounting for mean score shift, the gap yields a population lower bound on target calibration error, yet identical gaps can induce opposite transfer outcomes. Exact importance weighting recovers target proper loss under covariate shift, so failure of an estimated weighting pipeline does not by itself establish conditional shift. A finite-sample simultaneous lower certificate converts the population bound into a one-sided rejection rule using audit labels disjoint from evaluation outcomes. The leak-free gap correlation is 0.25 (95% CI [-0.09, 0.55]), falls to 0.09 on the second generator, and does not support a generator-invariant association. The certificate retains nominal coverage but has power 0.13 even at m=1024, whereas target-domain temperature scaling with 16 labels reaches harm rate 0.09, compared with 0.34 for source-fitted Platt scaling. Accuracy gaps are therefore weak warning signals for scalar probability transfer, not deployment certificates.

AI 总结：针对节点分类GNN，提出分离混合权重与logit定位误差的方法，通过精确质量线性规划和后处理修复，在八个数据集上提升准确率至65.3%，并发现平移主导增益。

链接：https://arxiv.org/abs/2609.27092

作者：Nadi Tomeh, Hugo Attali

英文摘要：Many node-classification GNNs apply a linear classifier to a nonnegative mixture of local messages. An error can reflect either poor mixture weights or a reachable logit set poorly positioned for the classifier. We separate these causes with an exact-mass linear program and two learned post-hoc repairs. Every reweighted prediction has an equivalent centered logit translation, but only translations in a message-induced displacement set are realizable by reweighting. Across eight datasets, eight GNN backbones, and ten splits, mean accuracy rises from 62.6% for the frozen models to 63.8% with reweighting and 65.3% with set-conditioned translation. A parameter-matched node-only translator reaches 64.6%, showing that translation explains most of the gain while the message set supplies a smaller additional benefit. Although oracle reweighting can correct many errors, label-free reweighting captures little of this potential: local evidence is often present but hard to select, and relaxing the evidence constraint is more effective than learning within it.

AI 总结：针对大规模图神经网络训练，提出基于电阻曲率的子图采样框架ERC-LG，利用Johnson-Lindenstrauss投影和正则化多GPU共轭梯度求解器高效近似曲率，在多数数据集上提升下游分类准确率。

链接：https://arxiv.org/abs/2609.27209

机构：South China Normal University（华南师范大学）； Academy of Mathematics and Systems Science, Chinese Academy of Sciences（中国科学院数学与系统科学研究院）

作者：Chaoqun Fei, Tinglve Zhou, Tianyong Hao, Yangyang Li

英文摘要： Subgraph sampling reduces the training cost of large-scale graph neural networks, but sampling criteria may overlook the geometric roles of edges. We propose a resistance-curvature-guided sampling framework built on ERC-LG, a curvature approximation method for large-scale graphs. ERC-LG combines Johnson-Lindenstrauss projections with regularized multi-GPU batched conjugate gradient solvers, avoiding explicit Laplacian pseudoinverse computation and full embedding storage. The resulting curvature informs node- and edge-sampling probabilities for constructing GNN training subgraphs. Experiments show numerical agreement with pseudoinverse-based curvature and reduced runtime compared with CG-only computation. ERC-LG-based sampling variants achieve the highest mean accuracy on six of seven real-world datasets in downstream node classification.

AI 总结：针对术后AKI早期预测，提出SynerT系列模型，结合波形与临床上下文，在严格泄漏感知框架下评估，堆叠集成SynerT-Stack取得最佳性能。

链接：https://arxiv.org/abs/2609.26848

机构：National Economics University（国民经济大学）

作者：Quang Minh Nguyen, Duc Minh Le, Ho Nhat Minh Nguyen, Thuy Quynh Nguyen, Trong Nghia Nguyen

英文摘要：Postoperative acute kidney injury (AKI) after major non-cardiac surgery carries substantial morbidity, yet early intraoperative risk stratification remains difficult. In this retrospective cohort study, we propose SynerT, a waveform-only hybrid temporal backbone that combines a causal dilated TCN with a hierarchy of dilated recurrent layers to encode early intraoperative physiologic trajectories for AKI risk prediction. Building on SynerT, we further design two model variants that extend the backbone with structured clinical context: SynerT-MM, a late-fusion multimodal extension that integrates hemodynamic burden summaries and preoperative covariates, and SynerTStack, a leakage-safe stacked ensemble that combines cross-validated predictions from SynerT-MM with strong tabular baselines at the meta-learning stage. All models are evaluated under a strict leakage-aware framework on VitalDB, a high-fidelity perioperative database, with prediction restricted to information available within the first 60 intraoperative minutes. Among 2,413 waveform-usable cases (180 AKI-positive; 7.46% prevalence), SynerT fell well below strong structured-data baselines, demonstrating that waveform-only temporal modeling is insufficient under strict early constraints. SynerTMM recovered discrimination by incorporating hemodynamic burden summaries and preoperative covariates, and SynerT-Stack achieved the best overall performance across AUROC, AUPRC, and F1-max. Cross-fitted Platt recalibration substantially corrected calibration defects in both multimodal variants, and decision-curve analysis confirmed the recalibrated stacked model delivered the strongest net clinical benefit across low-to-intermediate thresholds.

AI 总结：针对异构视觉骨干堆叠中的多重共线性与校准崩溃问题，提出CORE-STACK+预处理流程，含核冗余过滤、元特征门、谱自适应岭惩罚和贝叶斯混合器，在六个基准上显著提升精度与校准并降低计算开销。

链接：https://arxiv.org/abs/2609.26905

机构：Istanbul Technical University（伊斯坦布尔理工大学）

作者：Noor Islam S. Mohammad

英文摘要：Stacking heterogeneous vision backbones (CNNs, ViTs, and hybrids) is the de facto recipe for accuracy, calibration, and robustness, yet two coupled pathologies limit its returns. Prediction-space multicollinearity ill-conditions the meta-learner's Gram matrix, inflating weight variance and producing brittle solutions on a thin manifold. Calibration collapse compounds constituent miscalibration through naive linear stacking, so adding more models can hurt expected calibration error (ECE). Existing remedies, ridge regularization, greedy selection, model soups, and SWAG address at most one of these issues, and none jointly target conditioning and calibration in heterogeneous prediction pools. We introduce CORE-STACK+, a preconditioning pipeline with four components: (i) a kernelized redundancy filter that removes non-linear inter-model dependencies invisible to Pearson correlation, using Centered Kernel Alignment (CKA) [23]; (ii) a $<15$K-parameter differentiable meta-feature gate that learns per-sample attention over ensemble statistics; (iii) a spectrum-adaptive Ridge penalty $lambda^{star}=lmax(Chat)/SNR(Chat)$ derived from a Marchenko-Pastur signal-noise decomposition, eliminating nested cross-validation; and (iv) a Laplace-approximate Bayesian blender replacing inverse-RMSE heuristics. We prove a PAC-Bayes excess-risk bound that, for the first time, jointly accounts for prediction-space redundancy and meta-learner capacity. Across six benchmarks, CORE-STACK+ delivers $+1.8\%$ top-1 on ImageNet-1K, $-4.2$ mCE on ImageNet-C, $+0.9$ mIoU on ADE20K, and $+1.3$ AP on COCO, while reducing retained models by 35-57% and inference FLOPs by up to $41%$. ECE improves $2.1\times$ over deep ensembles without post hoc temperature scaling.

AI 总结：针对负荷削减导致数据稀缺的光伏预测难题，提出结合共形分位数回归的迁移学习框架，显著提升点预测精度与区间可靠性。

链接：https://arxiv.org/abs/2609.26959

机构：Green University of Bangladesh（孟加拉国绿色大学）

作者：Rakib Abdullah, K. M. Tahlil Mahfuz Faruk

英文摘要：Solar photovoltaic (PV) forecasting in regions affected by load shedding is challenging because reliable historical observations are scarce. This study proposes a transfer learning framework combined with Conformalized Quantile Regression (CQR) to improve PV power forecasting and provide reliable uncertainty estimates under severe data scarcity. A source-domain PV dataset from Alice Springs, Australia, is used to pretrain a temporal forecasting model, which is then adapted to simulated Bangladesh PV data representing different levels of historical availability. Experimental results show that transfer learning reduces RMSE by up to 23.7% when only one month of target-domain data is available and by 13.7% with three months of data. The proposed Transfer Learning plus CQR framework achieves 94.3% empirical coverage with three months of target data while producing prediction intervals that are 14% narrower than those obtained without transfer learning. These results demonstrate that combining transfer learning with conformal uncertainty quantification can improve both point forecasting accuracy and uncertainty reliability when target-domain PV data are severely limited.

AI 总结： 本文提出 ZO-COSMO，一种去中心化零阶优化方法，通过无索引一跳混合实现稀疏通信，在匹配负载下显著提升准确率，并给出收敛保证。

链接：https://arxiv.org/abs/2609.27199

机构：Hubei University（湖北大学）； Wuhan University（武汉大学）； Shanghai Jiao Tong University（上海交通大学）； Baidu（百度）

作者：Shengjun Zhang, Tingyi Liu, Heng Zhang, Dong Xie

英文摘要：Sparse communication in decentralized zeroth-order learning requires compatible peer-state coordinates. We characterize this one-hop condition and develop \textsf{ZO-COSMO}, coupling two-query estimation with average-preserving masked consensus using $q$ values per active link. Global supports serve all-neighbor mixing; matching updates require agreement only within each pair. We derive a sharp contraction-per-scalar bound within the matching class and convergence guarantees for the core and sparse-momentum updates. At fixed matching, exact moment identities characterize how shared directions preserve gradient-heterogeneity cancellation and redistribute estimation error and disagreement. Mechanism experiments cover unequal curvatures, noise, and sparse momentum. Further tests span $64$ synthetic agents and eight logical Qwen LoRA workers. At matched payload budgets, Qwen2-7B QNLI gains $3.65$ accuracy points over explicit-index Rand-$k$; edge-local updates gain $3.42$ and $2.53$ points over all-neighbor mixing on eight-worker complete and ring graphs. A matched-first-step ablation gives a $3.92$-point momentum benefit. Seed-aware and same-matching controls distinguish encoding, scheduling, and query correlation.

AI 总结：本文研究一类由块对角因子与固定置换构成的低参数正交矩阵，通过黎曼几何推导其流形结构，提出高效优化算法，并应用于矩阵逼近与大语言模型微调。

链接：https://arxiv.org/abs/2609.27982

机构：HSE University（高等经济大学）

作者：Ali Aliev, Maxim Rakhuba

英文摘要：In this paper, we are concerned with matrices formed by block-diagonal factors interleaved with fixed permutations -- a flexible family of structured matrices. This class has recently drawn interest in deep learning architectures for its balanced expressivity-efficiency trade-off, yet efficient computational strategies for working with it remain to be found. We approach this problem through Riemannian geometry and examine under what conditions this class admits a smooth manifold structure. For the practically important case of orthogonal two-factor matrices, we derive the essential Riemannian tools and propose efficient algorithms for their implementation. The algorithms leverage automatic differentiation, support parameter sharing within each factor, and avoid explicit dense matrix construction. We test them within the Riemannian optimization framework on the best matrix approximation problem and for parameter-efficient fine-tuning of large language models. Beyond the two-factor setting, we study the geometric and matrix-theoretic properties of factorizations with a larger number of block-diagonal factors.

AI 总结：PolyBench26是一个开放基准，包含近25万个聚合物性能数据点，支持四项评估任务，发现基于图的模型在性能预测中误差最低且稳健，为复杂聚合物设计提供可复现基础。

链接：https://arxiv.org/abs/2609.27036

作者：Robert W. Learsch, Nicholas Liesen, Daniel S. Levine, Anna M. Hiszpanski, Evan R. Antoniuk

英文摘要：Polymer property prediction lacks open, standardized benchmarks that enable rigorous comparison of machine-learning methods, with existing resources covering only a narrow fraction of polymer architectures, such as homopolymers. We introduce Polymer Benchmark 2026 (PolyBench26), an open dataset comprising nearly 250,000 polymer-property datapoints across eight physical properties, including data from experimental measurements, density functional theory, and molecular dynamics. The benchmark supports four evaluation tasks across homopolymers and alternating, random, and block copolymers: in-distribution property prediction, dataset-size scaling, repeat-unit complexity, and transfer to held-out polymer architectures. We compare language model, graph-based, and descriptor-based approaches and find graph-based models provide the lowest errors in property prediction, retain their advantage across the evaluated training-set sizes, and remain robust to increasing repeat-unit complexity. PolyBench26 provides a reproducible foundation for developing models for the increasingly complex polymer design space. The PolyBench26 benchmark is available open-source at this https URL.

AI 总结：本研究评估联邦可穿戴设备入职策略，发现个体级负迁移导致平均准确率掩盖失败，提供可审计基准而非普遍优越性声明。

链接：https://arxiv.org/abs/2609.27819

机构：Jamia Hamdard（贾米亚哈姆达德大学）； Homi Bhabha National Institute（霍米·巴巴国家研究所）； Variable Energy Cyclotron Centre（可变能量回旋加速器中心）； Gargi Memorial Institute of Technology（加尔吉纪念理工学院）； Maulana Abul Kalam Azad University of Technology（毛拉纳·阿布·卡拉姆·阿扎德技术大学）

作者：Rahil Aftab, Vineet Kumar Rakesh, Soumya Mazumdar, Tapas Samanta

英文摘要： Federated wearable models eventually serve people absent from source training, but favorable average accuracy does not establish that unlabeled onboarding helps each person. We evaluate six core onboarding strategies on five wearable datasets under a leakage-controlled protocol that fixes source checkpoints, estimates normalization from source data only, separates calibration from evaluation recordings, and performs inference over held-out people rather than windows, devices, or random seeds. Completing all eligible HHAR and PAMAP2 outer-person rotations materially changes the conclusion obtained from the original frozen fold. On HHAR, balanced accuracy on that single person is 95.6-97.2% across methods versus 78.3-83.0% over all nine users, a reduction of 13.8-17.8 percentage points (pp). The displayed mean leader changes on both datasets, while paired leader-runner bootstrap intervals include zero and do not resolve a superior method. No adaptive core mechanism combines positive mean gain in all five datasets with zero seed-averaged person-level losses greater than 2 percentage points (pp). FedBN has one such loss and ATP-style adaptation has eight; Feature-only has none after seed averaging, but its exact one-sided 95% upper bound is 7.6%. A complementary seed-person stress audit records 4, 22, and 10 harmful realizations out of 114 for FedBN, ATP-style, and Feature-only, respectively; these are repeated realizations, not independent participants. Tail quality, calibration availability, and fall-window specificity reveal additional failures hidden by mean accuracy. The study therefore provides an auditable development benchmark and failure map rather than a universal-superiority or deployment-safety claim.

AI 总结：Signal2Symbol通过神经符号框架将生理信号转为符号序列，结合稀有项集挖掘与Allen/FCA推理，实现可解释的时序异常检测与复合时间解释。

链接：https://arxiv.org/abs/2609.26820

机构：New York University（纽约大学）； Koc University（科奇大学）

作者：Naser Mansour, Sidahmed Benabderrahmane, Ameer Rahwan

英文摘要：Physiological time series such as electrocardiograms (ECG) and electroencephalograms (EEG) exhibit complex temporal structure, substantial acquisition variability, and a strong need for transparent decision-making. Although deep models can achieve high detection performance, they often provide limited insight into why a segment is anomalous, how local anomalies relate over time, and whether a detection belongs to a broader recurring pattern. We propose Signal2Symbol, a neuro-symbolic framework for explainable biosignal anomaly detection. The method first converts ECG/EEG signals into symbolic sequences using either a learned VQ-VAE (Vector Quantized Variational Autoencoder) codebook or a SAX (Symbolic Aggregate approXimation) baseline. It then constructs bigram enriched token-window transactions and scores anomalies through rare itemset evidence derived from minimal rare itemset mining. Detected anomalous windows are merged into intervals and related using Allen interval algebra, enabling composite temporal explanations such as escalation chains, artifact overlap, and cross-channel synchrony. Finally, we introduce a rare temporal concept lattice based on Formal Concept Analysis (FCA), which groups anomalous intervals by shared rare symbolic evidence, Allen temporal relations, channel context, and robustness attributes. The resulting Galois lattice compresses many local detections into interpretable families of temporal-symbolic anomalies. We evaluate on three public benchmarks: MIT-BIH Arrhythmia (beat-level ECG), PTB-XL (record-level ECG), and the Bonn EEG dataset (segment-level EEG). We stress-test robustness under additive noise and baseline-wander perturbations. The results highlight the value of neuro-symbolic tokenization for temporal anomaly analysis and show that Allen/FCA reasoning provides compact, interpretable summaries of local detections.

AI 总结：HARN是一种层次关联共振网络，通过持久表示和事件驱动更新实现多时间框架预测，在四个资产上达到与基线相当的误差，并验证了组件有效性。

链接：https://arxiv.org/abs/2609.26822

机构：Dr. APJ Abdul Kalam Technical University（APJ阿卜杜勒·卡拉姆技术大学）

作者：Nabeel Ahmad Saidd

英文摘要：Financial time series evolve across multiple temporal resolutions, challenging forecasting systems to incorporate newly available information without repeatedly recomputing unchanged representations. We introduce HARN, a Hierarchical Associative Resonance Network for event-driven multi-timeframe forecasting. HARN maintains persistent representations across temporal levels and updates each level only when its corresponding completed bar becomes available. The architecture combines causal multi-scale temporal encoding, gated associative memory, cross-level resonance, and hierarchical evidence aggregation, with forecasting performed in basis-point space and reconstructed to the original price scale. We evaluate HARN on four assets spanning equity, foreign exchange, and commodity markets using multiple random seeds and component ablations. HARN achieves competitive reconstructed-price forecasting errors against single-timeframe PatchTST and TimeXer baselines, while ablations reveal the effects of removing individual components across assets and timeframes. A code-level audit further examines consistency between the implementation and the defined event-driven causal protocol. The results position HARN as a persistent multi-timeframe forecasting framework rather than evidence of universal predictive superiority.

AI 总结：本文提出深度离散时间耗散递归神经网络（DissipNet），通过结构权重约束和专门训练算法显式强制耗散性，利用李雅普诺夫理论保证稳定性，并在建模应用中优于朴素RNN和PINN模型。

链接：https://arxiv.org/abs/2609.27186

机构：Sungkyunkwan University（成均馆大学）

作者：Tuan Luong, Hyungpil Moon

英文摘要：Physical AI has gained increasing attention for its role in developing AI systems that better understand, predict, and control real-world dynamics. Achieving this requires AI models that not only achieve high prediction accuracy but also preserve fundamental physical properties of dynamical systems. In this paper, we propose a deep discrete-time dissipative recurrent neural network (DissipNet) that explicitly enforces dissipativity, a key property related to stability and energy dissipation, through structural weight constraints and a dedicated training algorithm. By construction, the proposed network is capable of learning dissipative dynamics while preserving their inherent stability, which is formally analyzed using Lyapunov theory. In contrast to Physics-Informed Neural Networks (PINNs), which incorporate governing equations into the training loss but do not guarantee preservation of internal analytical properties such as dissipativity or passivity, our approach provides explicit guarantees on stability at the model level. We demonstrate the effectiveness of the proposed method through several modeling applications, and compare its performance with a naive recurrent neural network (RNN) and a PINN-based model.

AI 总结： 针对顺序推荐系统缺乏可解释性基准的问题，提出双模型掩码度量，系统评估十种XAI方法，发现梯度类方法最忠实，注意力权重需梯度加权才可靠。

链接：https://arxiv.org/abs/2609.27201

机构：Capital One（第一资本）

作者：Akash Pandey, Kanisha Shah, Addrish Roy, Dwipam Katariya, Hongyangyang Shi, Amanda Ding, Kalanand Mishra, Pranab Mohanty

英文摘要：Sequential RecSys are central to modern personalization, exploiting user's historical interaction sequences to drive next-step decisions. Deep learning models, particularly CNN and Transformer-based architectures, have proven highly effective at capturing temporal dependencies in these histories. For transparency and trust, understanding which past interactions drive a given recommendation is increasingly important --- both for developers auditing model behavior and for users seeking a rationale. However, the non-linearities that give these models their predictive power also render them black boxes, making it difficult to attribute decisions to specific interactions. While gradient-based, perturbation-based, and attention-based explainability methods exist, a systematic benchmark of their faithfulness for sequential recommendation is missing. We address this gap by introducing a dual-model masking metric in which one model supplies per-timestep attribution scores and a separately trained, masking-robust probe measures the resulting change in predicted probability. Using this metric, we benchmark ten XAI methods across CNN, Transformer, SASRec, and BERT4Rec backbones on KuaiRand and MovieLens, complemented by analyses of temporal attribution patterns, item popularity confounding, and robustness to input corruption. Our key findings are: (1) gradient-based methods, particularly GradientSHAP and Integrated Gradients, yield the most faithful and robust attributions; (2) raw attention weights are unreliable, but gradient-weighted attention restores faithfulness on shorter sequences, with degradation on longer horizons as softmax attention probabilities converge toward uniform importance scores, diminishing the method's ability to identify informative interactions; and (3) temporal attribution patterns in faithful methods reflect genuine task structure rather than recency or popularity bias.

AI 总结：SR-Fraud提出结果监督的反思式LLM框架，通过冻结智能体实时评分与离线反思验证假设，在非平稳支付欺诈检测中提升所有指标并捕捉新兴欺诈。

链接：https://arxiv.org/abs/2609.27287

机构：Coinbase, Inc.（Coinbase公司）； The Ohio State University（俄亥俄州立大学）

作者：Xuwei Tan, Yao Ma, Xueru Zhang

英文摘要：Real-time payment fraud detection is a non-stationary streaming prediction problem: adversaries adapt before supervised labels mature, and localized burst attacks can cause losses before retraining. Production systems typically rely on tabular classifiers and rules, which can struggle to capture these emerging sequential patterns before periodic retraining occurs. We present SR-Fraud, an outcome-supervised reflective LLM framework that decouples request-time decisions from offline adaptation. A frozen, stateless agent scores each transaction from a Hybrid Episodic Window to track behavioral shifts, while an offline reflection agent proposes boundary hypotheses from matured errors. A deterministic verifier then admits only supported hypotheses into an executable knowledge state. On a production payment-fraud benchmark, SR-Fraud improves all detection metrics over its frozen decision agent, obtains higher point estimates than static and periodically retrained CatBoost, and detects an emerging fraud burst.

AI 总结：针对异常数据稀缺导致模型选择困难的问题，提出将AUC上界作为可微的无异常目标，直接优化异常检测系统的连续参数（如集成权重和分数重缩放），实验表明该方法优于传统选择方法且对伪异常构造不敏感。

链接：https://arxiv.org/abs/2609.27362

机构：Aalborg University（奥尔堡大学）； Pioneer Centre for AI（先锋人工智能中心）

作者：Kevin Wilkinghoff, Zheng-Hua Tan

英文摘要：Anomalies are rare, and anomalous data are often unavailable during development, making it difficult to determine which anomaly detection models and configurations will generalize to unseen anomalies. Recent approaches address this challenge by generating pseudo-anomalies and using bounds on the achievable area under the ROC curve (AUC) to select the optimal configuration from a finite set of candidates. Instead, we use the AUC bound as a differentiable, anomaly-free objective for directly optimizing continuous parameters of anomaly detection systems. We demonstrate this framework by optimizing ensemble weights and introducing a learnable score-rescaling mechanism that adapts pseudo-anomaly scores, enabling optimization beyond a predefined candidate set. Experiments across multiple datasets and embedding models show that AUC-bound optimization achieves significant performance gains over conventional model selection and prior development-set-based parameter selection. The results further show that direct optimization is less sensitive to the choice of pseudo-anomaly construction.

AI 总结：FWBench通过1,251个案例评估语言模型在预算约束下选择和使用时间序列预测工具进行决策的能力，发现GPT-6 Astra以2.5%预算实现优于固定策略的决策质量。

链接：https://arxiv.org/abs/2609.27385

机构：Neurogica Inc.（Neurogica 公司）

作者：Shunya Nagashima

英文摘要：Time-series foundation models (TSFMs) provide forecasts for operational decisions, but accuracy alone does not determine their value. Evaluating agents that use these models requires measuring decision quality and forecast cost. FWBench evaluates this capability on 1,251 electricity and cycle-hire cases using fixed forecast tools and simulated capacity contracts. Agents select models, histories and horizons, then submit capacities to minimize a stated loss-cost objective. We evaluated two hosted and eight local configurations, including small language models, and tested local models with and without TSFMs. GPT-6 Astra bought inexpensive short-horizon forecasts selectively, using 2.5% of the budget, and outperformed fixed policies when the saved decisions were scored with three loss-cost weightings. FWBench enables reproducible evaluation of how language models select and use time-series forecasts to make decisions under cost constraints.

AI 总结：本文提出ROOSTER模块，通过可学习的周期梳状偏置统一处理时间序列预测与PPG到生命体征重建，在多个基准上取得最优性能，并验证了相对对齐机制的有效性。

链接：https://arxiv.org/abs/2609.27473

机构：Neurogica Inc.（Neurogica 公司）

作者：Ragamayi Puli, Shunya Nagashima

英文摘要：PPG-to-vital-sign reconstruction turns a wrist-worn photoplethysmogram into clinical waveforms such as the ECG. Long-horizon multivariate time-series forecasting underpins planning in energy, weather, and traffic. Both generate a target sequence from a condition sequence, and current models hard-code where each target position reads it, as a same-position copy or seasonal recurrence, so neither transfers between tasks. We propose ROOSTER, one conditioning module that handles vital-sign reconstruction and time-series forecasting alike by learning this correspondence. Its core is a periodic-comb bias over the target-condition offset whose center, period, and sharpness are learned per head, so one module settles on the identity alignment or a seasonal lag and reports which it found. On vital-sign reconstruction from PPG, ROOSTER outperformed the published baselines on four heart-rate and respiratory-rate benchmarks. On multivariate time-series forecasting, it achieved the best horizon-averaged MSE on four benchmarks and outperformed the forecasting model it extends on 20 of 24 dataset-horizon settings under matched three-seed training. An ablation study indicated that the relative bias, not content matching, carried the alignment.

AI 总结：提出LoReST网络，通过关系感知局部聚合和跨区域注意力，高效建模大规模交通预测中的局部异质性与长距离上下文，在LargeST基准上显著降低误差。

链接：https://arxiv.org/abs/2609.27637

机构：Northwestern Polytechnical University（西北工业大学）； City University of Hong Kong（香港城市大学）； Shenzhen Research Institute, City University of Hong Kong（香港城市大学深圳研究院）； Northwest University（西北大学）

作者：Qi Feng, Zidong Wang, Bo Li, Xiaoguang Gao, Jiayu Zhang, Chenfeng Wang, Kaifang Wan

英文摘要：Traffic flow forecasting is essential to intelligent transportation systems. Large-scale traffic forecasting requires jointly modeling local spatial dependencies and cross-region this http URL dependencies between geographically neighboring nodes are heterogeneous due to differences in road identity and travel direction, while acquiring global information through allpairs node interactions incurs substantial computational costs. Therefore, capturing local heterogeneity while efficiently acquiring long-range context remains an important challenge in largescale traffic forecasting. To address these challenges, we propose LoReST, a Local-Region Spatial Temporal network that models spatial dependencies at two complementary granularities: node neighborhoods and road network regions. Specifically, relation-aware local aggregation captures heterogeneous dependencies within geographic neighborhoods through road and direction specific feature transformations. Cross-region interaction constructs region representations through mean pooling, exchanges long range context via inter-region attention, and broadcasts it back to nodes. By integrating local information aggregation with crossregion interaction, LoReST is able to effectively achieve spatial dependency learning in large-scale road networks. Experiments on four datasets of the LargeST benchmark show average relative reductions of 4.78%, 3.60%, and 5.75% in MAE, RMSE, and MAPE, respectively.

AI 总结：MENO是一种基于流形函数编码器的内存高效神经算子，用于求解偏微分方程，具有内存占用小、支持任意几何与离散化、泛化能力强等优势，在多数基准上达到最优精度。

链接：https://arxiv.org/abs/2609.27739

机构：Peking University（北京大学）； National Engineering Laboratory for Big Data Analysis and Applications, Peking University（北京大学大数据分析与应用技术国家工程实验室）； Chongqing Research Institute of Big Data, Peking University（北京大学重庆大数据研究院）

作者：Shengyang Xu, Weijun Zhang, Jun Hu, Pengzhan Jin

英文摘要：We propose the Memory-Efficient Neural Operator (MENO) as a high-performance PDE neural solver based on the Manifold Function Encoder (MFE). MENO features three primary advantages: (1) MENO has a significantly smaller memory footprint and much faster training speed than other popular architectures, with the memory footprint being independent of the data resolution, and therefore holds the potential for scaling up to large-scale models. (2) MENO can accept PDE inputs of arbitrary form, including arbitrary geometric domains and arbitrary discretizations. In particular, it is capable of handling cross-geometry scenarios, i.e., where the input functions and the output solutions are defined on different manifolds. (3) MENO exhibits strong generalization capability, and achieves the best accuracy on most of the benchmarks we tested, compared with the results reported in the literature. The code is available on GitHub at this https URL, and all numerical examples in this paper can be run with a single command to reproduce the reported results.

AI 总结：本研究通过在生产市场面板上对24种预测方法进行基准测试，证明评估设计选择（如分析单位、误差汇总和评分方式）能逆转或消解预测排行榜结论，并量化了部署系统选择规则的效果。

链接：https://arxiv.org/abs/2609.27867

机构：Field Nation LLC（Field Nation有限责任公司）

作者：Md Rezwanul Islam, Wael Mohammed

英文摘要： A forecasting benchmark reports which method won. We show that the answer is set by the evaluator's choices before any model is fitted. We benchmark 24 forecasting methods and one textbook reference, including six 2025-era time series foundation models, on a production marketplace panel of 1,887 business customers over 67 months. We hold the data, the horizon and the period fixed, and vary only the evaluation design. Three choices each reverse or dissolve a headline conclusion. Changing the unit of analysis from the market total to the individual customer moves our production baseline from second of nineteen, beaten by nothing, to twenty-third of twenty-five. Nineteen of its twenty-four challengers beat it there. Changing how much error is pooled decides whether a Diebold-Mariano test finds anything at all. Scoring prediction intervals rather than point forecasts reorders the field almost completely, with a rank correlation of 0.02 on intermittent demand. We then measure what the deployed system gets from this. Its selection rule captures 55% of the distance between doing nothing and choosing with hindsight. The reversal is not a quirk of our data. We ran the released protocol, unchanged, on the public M5 retail panel. The same baseline shape places first at the market total and last per series, beaten by everything, and a replayed selection rule closes 64.7% of the same floor-to-ceiling distance there. Adding five zero-shot foundation models to that roster changes who wins at the total, not the shape. The bands' blind spot travels too: conformal bands under-cover most on the spikiest items. Splitting our own panel into ever smaller groups turns the contrast into a curve: the baseline's rank worsens at every level of disaggregation. We release the evaluation protocol and report an error of our own that inverted a result before we caught it.

AI 总结：PISCES是一种物理信息卷积自编码器，在无标签的OMNI太阳风数据上训练，通过分解异常分数并利用物理约束，实现L1点太阳风瞬变的早期检测与预警，可提前于急始事件。

链接：https://arxiv.org/abs/2609.28022

机构：UCLA（加州大学洛杉矶分校）； University of Colorado（科罗拉多大学）； NASA Jet Propulsion Laboratory（美国国家航空航天局喷气推进实验室）

作者：Kevin Lee, Alison J. March

英文摘要：Space weather early warning depends on detecting solar wind transients in in-situ measurements at the first Sun-Earth Lagrange point (L1), before they reach Earth. Fixed thresholds can miss combined magnetic and plasma structure, and many learning methods provide a single anomaly score. We present the Physics-Informed Solar-wind Convolutional autoEncoder for Space-weather (PISCES), a convolutional autoencoder trained without catalog labels on OMNI solar wind measurements under physics constraints. Its loss includes magnetic field consistency, an empirical relation between temperature and velocity, the Parker spiral angle, and penalties on changes between consecutive one-minute samples in derived quantities calculated from the reconstruction. At inference, PISCES separates the anomaly score into magnetic and plasma reconstruction errors, physics relations, and residual corrections, and reports the magnitude of each contribution. Attenuation of the skip connections, selected on validation data, improves average precision for the trained models, while the untrained scores remain nearly the same. The trained models also give a more consistent ordering of these physical contributions. After smoothing with a trailing median, the alarms can precede independently observed sudden commencements, including positive sudden impulses.

AI 总结：本研究分析Terminal-Bench 3/Frontier-Bench 0.1的生产记录，提出区分真实难度与虚假难度的方法，发现仅125个全失败任务中78个为认证未解决，强调基准应报告全失败任务证据。

链接：https://arxiv.org/abs/2609.26826

作者：Edward Lue Chee Lip, Boden Moraski, Tim Knappe, Lang Xiong, Sarvesh Gharat, Antonio Mari, Ivan Bercovich

英文摘要：Frontier benchmarks need tasks that current models cannot solve. But a task that no model solves is not automatically a hard task. The same zero pass rate can come from a real capability gap, but it can also come from missing context, a broken reference solution, infrastructure failure, or a verifier that can be bypassed. In this paper, we study this issue using a frozen Terminal-Bench 3 / Frontier-Bench 0.1 production record with 1,081 pull requests, 639 scored tasks, 28,801 trials, and $105,933 in logged agent spend. We ask what an all-fail task actually certifies. For the 125 tasks with no honest pass, we combine task artifacts, reference-solution runs, empty-solution controls, adversarial trials, trajectories, telemetry, and review records, and apply an ordered validity screen. Only 78 of the 125 tasks survive as certified-unsolved candidates. The remaining tasks include 14 with broken oracles, 8 dominated by infrastructure failures, 4 that are only passable through verifier bypasses, and 21 whose solvability is not certified by the available evidence. Thus, lack of saturation and genuine difficulty are not the same thing. The certified-unsolved label is also narrow: it means that the authored route passed, infrastructure did not dominate, no strict bypass was observed, and all evaluated agents failed. It does not prove intrinsic hardness, verifier completeness, or failure at the intended capability. We further analyze rejected submissions and passing tasks to show that pass rate alone cannot explain why a task is difficult. Overall, our results suggest that frontier benchmarks should report the evidence behind their all-fail tasks before using them as capability claims.

AI 总结：COPE通过为每位用户分配可学习嵌入并结合自我评估生成代理奖励，在稀疏反馈下实现LLM的持续个性化优化，实验证明其优于现有基线并与检索增强提示互补。

链接：https://arxiv.org/abs/2609.26853

机构：University of Science and Technology of China（中国科学技术大学）； Qwen Business Unit of Alibaba（阿里巴巴千问事业部）

作者：Ruike Cao, Fugen Yao, Liang Dong, Jian Xu, Guanjun Jiang, Li Xiao

英文摘要： While Large Language Models (LLMs) have achieved remarkable results across various benchmarks, their alignment with normative values often results in homogenized responses that fail to address diverse user preferences. Existing training-free methods often occupy valuable context windows through prompt engineering, while training-based methods typically remain static post-training, failing to support the continual optimization required in real-world settings. To address these challenges, we propose COPE (Continual Optimization with Personalized embedding and self-Evaluation), a novel optimization framework tailored for real-world-motivated interaction settings with sparse user feedback. Our framework assigns learnable personalized embeddings to each user and synergistically integrates preference capture, self-evaluation calibration, and personalized response optimization within a single update step. A key innovation of our method is the use of self-evaluation to generate proxy rewards, enabling continuous model updates even when explicit user feedback is unavailable. Experiments show that COPE consistently outperforms strong training-free and training-based baselines under sparse feedback, and remains complementary to Retrieval-Augmented Prompting (RAP). Further analyses confirm COPE's reliable self-evaluation, meaningful preference patterns, stable general capabilities, and robustness under shifting preferences and alternative evaluators.

AI 总结：本文证明在双动作模型中，即使工具缓存保持边际奖励分布，共享随机结果也可能逆转组归一化策略更新，表明边际有效性不足以保证训练等价性。

链接：https://arxiv.org/abs/2609.26866

作者：Shivam Gupta

英文摘要：Tool-result caching reduces repeated execution in agent training, but also couples rollout randomness. We study a two-action model in which independent and shared execution preserve every rollout's conditional reward distribution. Despite this marginal agreement, sharing one stochastic result per group can reverse the expected group-normalized policy update. We derive an exact finite-group expression: against a constant alternative, the shared update follows the probability of winning minus the probability of losing, rather than the difference in expected reward. A Bernoulli specialization yields a wrong-direction region and a non-vanishing update-variance floor as group size grows. Centering without group standard-deviation scaling preserves the expected-return direction in this model, using an existing estimator control. Exhaustive finite sums verify 540 configurations and 3,240 estimator evaluations, with a separate ordered-sequence checker. An implementation audit reproduces the sharing path in a pinned, unmodified TVCache stack using 256 scripted rollouts. These results do not measure language-model training performance or refute TVCache's deterministic-output contract. They establish that marginal output validity alone cannot certify a stochastic cache as training-equivalent.

AI 总结：提出PR-Smoother，一种保模拟器摊销平滑器，通过仅学习未来条件修正保持指定模拟器显式，实现非高斯后验平滑，并在Lorenz-96及Kolmogorov流中验证多模态捕获与高维扩展能力。

链接：https://arxiv.org/abs/2609.26890

机构：iTHEMS； RIKEN（理化学研究所）

作者：Yuta Tarumi

英文摘要：Many physical data assimilation (DA) workflows require smoothing methods that represent non-Gaussian posteriors over physical state variables, scale to high-dimensional simulators, train from observation windows alone, and remain compatible with calibration of the prescribed simulator. We introduce PR-Smoother, a simulator-preserving amortized smoother designed for this prescribed-simulator DA regime. Its key design principle is to keep the prescribed simulator explicit in both the evidence lower bound and the variational family: rather than learning replacement dynamics or a learned trajectory prior, PR-Smoother learns only future-conditioned corrections around the prescribed rollout. This yields an explicit non-Gaussian smoothing distribution over physical trajectories and supports joint state, parameter, and sensor-bias learning from observations alone. The variational family contains the exact smoother in deterministic and linear-Gaussian limits. Empirically, PR-Smoother captures multimodal posteriors in 4-dimensional Lorenz-96, remains accurate under ambiguous nonlinear observations and process noise in 40-dimensional Lorenz-96, and scales to joint state-parameter-bias inference in 16,384-dimensional Kolmogorov flow.

AI 总结：本研究提出FAPE框架，通过八项跨领域评估发现后处理公平性约束在高差异场景下有效但可能恶化接近公平场景，主张采用基线差异筛查与持续监控替代单一部署时审计。

链接：https://arxiv.org/abs/2609.26955

作者：Nithin Raghava Ramachandra Narla

英文摘要：Fairness audits in production ML typically occur once, at deployment, on a single domain. Both fail in practice: fairness can shift after retraining or a changing user base, and interventions validated on one dataset are rarely tested across the heterogeneous domains an organization deploys. We present FAPE (Fairness Auditing for Production Environments), a four-stage framework evaluating a single post-processing intervention, Fairlearn's ThresholdOptimizer, across eight domain evaluations: criminal justice, income prediction, legal admissions, credit lending, agricultural lending, a multi-domain benchmark corpus, healthcare, and education. Each is scored on demographic parity and equalized odds difference, plus disparate impact ratio and accuracy cost where computable. Intervention effectiveness tracks baseline disparity magnitude: across model-domain pairs the constraint improved disparity in 9 of 14 high-disparity cases and worsened it in 3 of 4 near-fair ones. Each of the five high-disparity exceptions reverses under one of two measurement checks, a minimum group size or thresholds fit on held-out data. A CUSUM monitor started at deployment, tested on a simulated shift, separates constrained models that never met a 0.1 parity convention from those that met it and later regressed. A single deployment-time audit is therefore an unreliable guide, which argues for baseline-disparity screening and continuous monitoring

AI 总结：针对不平衡表格数据的大规模训练成本问题，提出线性时间核心集方法CRISP，通过重要性分层分配负类预算并加权样本，在减少93.2%训练行时保持99.7%精度，显著优于现有方法。

链接：https://arxiv.org/abs/2609.26962

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

作者：Hardhik Mohanty, Indrayana Rustandi, Mohamadreza Sheibani

英文摘要： Large imbalanced tabular datasets make repeated gradient-boosted tree training expensive. Existing coreset methods often lose accuracy when most majority examples are removed. We present CRISP (Coreset Reduction via Importance-Stratified Pruning), a linear-time method that allocates a negative-class budget across quantile strata of a proxy-model score. Sample weights account for unequal inclusion probabilities. At 95% negative-class reduction on a production fraud dataset, CRISP trains on approximately 1.70M of 25M rows and retains 99.7% of full-data Average Precision. This is a 93.2% reduction in total training rows. On public CriteoPrivateAds, CRISP has the highest mean Average Precision at each tested rate from 90% to 99.4% majority reduction. Sparkov results are mixed at lower rates, but CRISP has the highest mean at 99.2% and 99.4%. Ablations identify budget allocation and inverse-propensity weighting as the main sources of the production-dataset gain.

AI 总结：提出Lie-Taylor射流匹配的免求解器训练框架，直接在微控制器上拟合UDEs，通过噪声自适应机制在精度上匹配或超越传统基线，实现实时设备上训练。

链接：https://arxiv.org/abs/2609.26972

机构：The University of Texas at Austin（德克萨斯大学奥斯汀分校）

作者：Pranavanath Balamurali, Hrishi Kamireddy

英文摘要：Training Universal Differential Equations (UDEs) traditionally relies on backpropagating through numerical ODE solvers, creating memory footprints far exceeding the capabilities of edge microcontrollers. We present Lie-Taylor jet matching, a solver-free training framework that fits a hybrid vector field directly to the first and second time-derivatives of observed system states. These derivatives, the truncated Lie-Taylor jet, are estimated online via Savitzky-Golay filtering, yielding fully analytic gradients without automatic differentiation software. We evaluate whether eliminating the solver compromises accuracy against a conventional baseline (fixed-step RK4 integration, multiple shooting, exact discrete adjoints, Adam) sharing identical dynamics, noise models, network architectures, and metrics. While naive derivative matching degrades under sensor noise, our noise-adaptive mechanisms close and reverse this gap: full-rate phase-shifted sampling, a reservoir buffer, cosine-annealed optimization with weight averaging, on-device noise estimation, and polynomial-misfit quality gating. On a damped pendulum and chaotic double pendulum, our method matches or exceeds baseline accuracy at matched data windows and recovers unmodeled damping coefficients. Across noise levels from 0% to 5%, it attains a geometric-mean relative field error of 0.65x that of the baseline within 108 kB of static memory, compared with megabytes of solver tape. On an ESP32 microcontroller, the on-device run reaches a field error of 0.0020 and recovers the damping coefficient to c = 0.400 (true 0.400) within 61.3 kB of static memory and 7.24 ms per update (18.1% duty cycle at 25 Hz), confirming real-time on-device training is feasible without a numerical solver.

AI 总结：GeoRVQ提出解码器感知的几何目标用于生理信号残差令牌预测，通过粗到细掩码建模提升波形和事件保留，在多个数据集上显著改善解码距离和R峰F1。

链接：https://arxiv.org/abs/2609.27018

机构：University of Twente（特文特大学）

作者：Bo Cui, Yaowen Zhang

英文摘要：Residual vector quantization (RVQ) turns physiological waveforms into compact token sequences, but conventional masked modeling treats every incorrect token as equally costly. We propose GeoRVQ, a coarse-to-fine masked token model whose objective reflects the local response of a frozen waveform decoder. Decoder-induced costs define geometry-aware soft targets and expected distortion, while quantizer-causal prediction follows residual dependencies from coarse to fine levels. In a descriptive aggregate over MIMIC-IV Waveform, VitalDB, and CODE-15\%, GeoRVQ increases exact token accuracy from $.133\pm.004$ to $.143\pm.003$, reduces decoded distance from $.606\pm.006$ to $.393\pm.007$, and increases R-peak F1 from $.784\pm.004$ to $.837\pm.008$ under matched model and training conditions. Across 45 held-out code substitutions, decoder-induced cost has a Spearman correlation of $.85$ with realized decoded cost, compared with $.54$ for Euclidean codeword distance. These results indicate that decoder-aware objectives can improve waveform and event preservation without requiring a large increase in exact token accuracy.

AI 总结：提出ChipMEM，一种面向EDA智能体的验证接地记忆层，结合跨任务程序性记忆与贝叶斯统计引导，仅在通过综合、仿真或形式化检查后存储技能，提升RTL优化与测试平台生成性能，并实现技能向未见任务的迁移。

链接：https://arxiv.org/abs/2609.27067

作者：Abdulrahman AlRabah, Joshua Mabry, Dilek Hakkani-Tür, Abdussalam Alawini, Hamid Shojaei, Kartik Hegde, Sandesh Adhikary

英文摘要：Large language model (LLM)-based agents use Electronic Design Automation (EDA) tools to generate and revise register-transfer-level (RTL) designs under synthesis and verification feedback. Recent methods learn from this feedback by distilling reusable skills from execution traces or by training on rewards derived from EDA-tools. Both methods are typically evaluated on the tasks that produced the experience. Repeated access to benchmark feedback on the same task can reward task-specific revision rather than creating reusable knowledge that transfers. We introduce ChipMEM, a verification-grounded memory layer for EDA agents. It combines cross-task procedural memory with within-trajectory statistical guidance. Its procedural component distills and stores a skill only after it passes synthesis, simulation, or formal checks, rather than relying on model self-assessments. A Bayesian component maintains hierarchical Beta estimates over tool-call outcomes and ranks recovery strategies that succeeded under comparable errors. A common adapter applies the same memory interface to RTL optimization and testbench-generation agents while preserving each domain's tools and acceptance criteria. We measure performance on training tasks and evaluate whether learned skills transfer to unseen tasks. On RTLRewriter-Bench, under matched model and tool settings, ChipMEM produces equivalence-passing outputs on 39/54 scored designs versus 35/54 without memory; on the 49-design short suite, mean area improvement is 8.69% versus 5.66%. On held-out CVDP tasks, ChipMEM with a frozen procedural library achieves 20/20 accepted outcomes versus 18/20 without memory in a single evaluation per setting.

AI 总结：针对未观测混杂导致风险评分低估高风险个体的问题，提出结合敏感性分析与Wasserstein分布鲁棒优化的IPW改进方法，在半合成数据上校准性能提升达29.2%。

链接：https://arxiv.org/abs/2609.27144

机构：University of Southern California（南加州大学）； University of Luxembourg（卢森堡大学）

作者：Ryan Edmonds, Yingxiao Ye, Sina Aghaei, Andrés Gómez, Çağıl Koçyiğit, Phebe Vayanos

英文摘要：We consider the problem of learning risk scores to prioritize individuals for scarce resources or interventions, from historical observational data affected by unobserved confounding. Decisions about who receives scarce resources are often guided by risk scores based on recorded characteristics, such as responses to a survey. These risk scores are increasingly being learned directly from observational data: historical records of individuals' characteristics, allocation decisions, and outcomes. Standard methods such as inverse propensity weighting (IPW), which corrects for the bias introduced by the historical allocation policy, can be used to learn accurate risk scores if the historical decision process is fully explained by the recorded characteristics. In practice, however, historical decisions often depend on unrecorded information, causing learned risk scores to systematically under-prioritize exactly the individuals whose unrecorded circumstances drove past prioritization. We propose a method for learning risk scores that are robust to this kind of unobserved confounding, building on IPW. Since propensity weights cannot be reliably estimated under unobserved confounding, we instead treat them as belonging to an uncertainty set determined by the observable data and domain-informed estimates of the degree of confounding, combining sensitivity analysis from causal inference with Wasserstein distributionally robust optimization. The resulting robust risk score learning problem admits a sample-based approximation that we reformulate as an exponential cone program compatible with off-the-shelf solvers. We demonstrate the effectiveness of our approach on semi-synthetic data derived from datasets in the UCI Machine Learning Repository. Our method improves calibration by up to 29.2% over traditional benchmarks and up to 11.1% over the state of the art, without compromising other metrics.

AI 总结：本文指出线性表示假说实为基于表示等价性的论断族，通过群作用形式化表示对象、生成过程与断言属性，并用于审计常见表示量与可解释性分析。

链接：https://arxiv.org/abs/2609.27158

机构：The Chinese University of Hong Kong（香港中文大学）

作者：Louie Hong Yao, Yuhao Li, Shengchao Liu

英文摘要：To make claims about representations that generalize beyond a particular trained model, we need to specify when two representations should count as equivalent. The Linear Representation Hypothesis is often discussed without making this equivalence explicit. Different notions of equivalence preserve different structures, so metrics, probes, and interventions that appear to study the same representation may in fact correspond to different hypotheses. We therefore argue that the Linear Representation Hypothesis is not one hypothesis but a family of claims distinguished by representation equivalence. We formalize this idea using group actions, specifying the representation object, the procedure that produces it, and the property ultimately asserted, while accounting for equivalences imposed by the model architecture. This framework clarifies how assumptions can change across metrics, reading points, and analysis stages, and we use it to audit common representation quantities and recent interpretability analyses.

AI 总结：本研究通过层次贝叶斯模型分析DeepSeek-R1-Distill模型，发现推理能力随规模收益递减，而效率不随规模提升，揭示朴素缩放的局限性。

链接：https://arxiv.org/abs/2609.27166

机构：Northeastern University（东北大学）； Complexity Science Hub Vienna（维也纳复杂性科学中心）； Santa Fe Institute（圣塔菲研究所）

作者：Moritz Laber, Zohair Shafi, Germans Savcisens, Brennan Klein, Matteo Chinazzi, Samuel V. Scarpino, Albert-László Barabási, Tina Eliassi-Rad

英文摘要：Capability and efficiency are two key dimensions of reasoning in large language models (LLMs). Capability refers to the ability to solve a given problem correctly, whereas efficiency refers to the ability to do so with limited resources. When LLMs use Chain-of-Thought (CoT) reasoning to solve problems of controlled hardness, both the number of problems solved correctly and the number of tokens required to reach a correct answer depend on problem hardness and model size. However, how these factors jointly shape capability and efficiency remains poorly understood. Here, we use hierarchical Bayesian models to evaluate the capability and efficiency of LLMs from the DeepSeek-R1-Distill model family across four classes of arithmetic and algorithmic reasoning problems. At a fixed model size, the probability of correctly solving an instance decays approximately exponentially with instance size, our proxy for problem hardness. The decay scale grows sublinearly with model size, indicating that larger models are more capable, but that capability gains diminish with scale. Output length grows as a power law with instance size, which serves as a proxy for difficulty. However, the parameters of this power law do not vary systematically with model size, suggesting that larger models do not become more efficient. Together, these findings reveal potential limitations of naive scaling as a strategy for developing more capable AI systems: capability improves with diminishing returns, while efficiency shows little to no improvement.

AI 总结：针对AI虚拟细胞中智能体模型对输入使用声明的不可靠问题，提出CELLAUDIT审计方法，通过源代码检查和预测贡献验证，结合证伪修订，提升模型性能与输入使用可靠性。

链接：https://arxiv.org/abs/2609.27234

作者：Mengran Li, Bo Li, Chengyang Zhang, Yang Yan, Jinfeng Xu, Zhenchao Tang

英文摘要： AI virtual cells aim to predict cellular responses to specified interventions, yet held-out predictive performance alone does not establish use of the supplied perturbation information. This prediction-claim gap matters in agentic model discovery, where language-model agents generate and revise predictors using score-based feedback. We introduce CELLAUDIT, which audits input-use claims by asking whether an input can enter the cited computation, whether fitted predictions depend on it, and whether that dependence improves prediction of observed response. On a paired morphology-transcriptomics perturbation benchmark (BBBC047), an agent-selected predictor attains a mean held-out Global Pearson correlation coefficient (PCC) of 0.3153 but remains invariant to compound replacement; a control-profile-only predictor reaches 0.3142. Source inspection identifies a compound-query pathway blocked by singleton key-value attention, and the invariance persists after refitting with disjoint control wells. In a stratified audit of 48 candidates across two linked tasks, 47 change predictions under compound replacement on both held-out folds, but only 20 show target-loss gains with intervals above zero on both folds. On BBBC047, falsification-guided revisions recover positive mean compound contributions while retaining gains over the control-profile-only baseline. In matched sci-Plex searches, audit-enriched feedback yields higher held-out performance and larger mean compound and dose contributions across five trajectories, although paired intervals span zero. Refitting fixed designs on an independently acquired cohort shows predictive generalization need not imply generalization of input-use claims: dose contribution persists, whereas support for compound identity does not. CELLAUDIT adds a falsification layer to agentic model discovery, moving from generate-score-revise toward discover-falsify-revise.

AI 总结：该研究挑战柏拉图式表征假说的局部收敛解释，提出结构尺度与比较对象的混淆，构建2×2框架评估关系结构与度量几何，发现关系结构在局部和全局尺度均稳健收敛，而度量几何收敛显著较弱。

链接：https://arxiv.org/abs/2609.27252

机构：KAIST（韩国科学技术院）； POSTECH（浦项科技大学）

作者：Junwon You, Mihyun Jang, Sangwoo Mo, Jae-Hun Jung

英文摘要：The Platonic Representation Hypothesis suggests that increasingly capable models converge toward shared representations. Recent work narrows this claim to shared local neighborhood relationships, finding that capacity-dependent trends in several global similarity measures largely disappear after calibration. We challenge this interpretation by showing that prior local-global comparisons confound structural scale (local versus global) with what is compared: relational structure, defined by which samples are related, versus metric geometry, characterized by quantitative relations such as distances, similarities, or correlations. To disentangle these factors, we construct a controlled $2\times2$ framework that evaluates both relational structure and metric geometry at local and global scales. We introduce $H_0$ skeleton overlap as a global counterpart to mutual $k$-nearest neighbors, together with matched distance-aware variants. Across vision-language models, relational structure exhibits robust representational convergence at both scales after calibration, whereas increasingly stringent distance agreement substantially weakens alignment and progressively flattens the capacity-dependent trend. We further extend the analysis beyond ambient Euclidean geometry by evaluating distance agreement under a Riemannian metric approximation and recover the same structure-geometry pattern. The pattern is also reproduced in video-text representations. Together, these results show that relational convergence extends beyond local neighborhoods to global spanning structure, whereas metric geometry exhibits substantially weaker convergence.

AI 总结：本工作提出Live Assistant框架，将直播协助形式化为是否、何时、向谁及沟通什么四个决策，通过自回归策略和两阶段训练，在真实直播数据上实现高准确率，建立选择性参与的协助范式。

链接：https://arxiv.org/abs/2609.27303

机构：Institute of Trustworthy Embodied AI, Fudan University（复旦大学可信具身智能研究院）； ByteDance TikTok（字节跳动TikTok）； Shanghai Innovation Institution（上海创新研究院）； Zhejiang University（浙江大学）

作者：Shujian Gao, Jiamei Yan, Yuchen Yang, Penghao Zhou, Qinglei Wang, Tiehan Fan, Yuan Wang, Zuxuan Wu, Yu-gang Jiang

英文摘要：Livestreams are long-lasting interactive environments where audiovisual content, viewer activity, host behavior, and platform signals evolve together, creating assistance needs that emerge from the stream itself. We introduce \liveassistant, a framework for mixed-initiative, role-conditioned assistance that formulates livestream interaction as four coupled decisions: \textbf{whether to act, when to act, whom to address, and what to communicate}. At each 10-second interval, one autoregressive policy consumes native audio and video with synchronized comments, gifts, viewer dynamics, and room metadata, then selects \textsc{OBS}, \textsc{MEM}, or \textsc{ANS}. \textsc{OBS} remains silent, \textsc{MEM} records a private semantic update, and \textsc{ANS} specifies a recipient, task, and grounded message. To support this task, we build a trajectory engine that reconstructs real livestream sessions into structured causal supervision, yielding over 320 hours of optimization trajectories and a human-reviewed benchmark of 275 clips and 13,812 decision intervals. We train the policy with Marker-Aware Multiturn Supervised Fine-Tuning (MA-MSFT), which strengthens sparse structured decisions, followed by Streaming Multiturn GSPO (SM-GSPO), which optimizes self-generated trajectories with turn- and trajectory-level credit. On the held-out benchmark, \liveassistant reaches 71.14 state accuracy, 72.67 recipient accuracy, and 58.41 task accuracy, with consistent gains over representative streaming and general multimodal baselines. Together, the formulation, benchmark, and training framework establish livestream assistance as selective participation in a shared social stream.

AI 总结：提出量化鲁棒去学习框架，通过曲率准则定位敏感权重，采用敏感度引导噪声正则化和遗忘关键优化，在MUSE和TOFU基准上实现量化弹性遗忘并保持模型效用。

链接：https://arxiv.org/abs/2609.27355

机构：University of North Carolina at Chapel Hill（北卡罗来纳大学教堂山分校）； University of Pittsburgh（匹兹堡大学）； University of Arizona（亚利桑那大学）； University of Central Florida（中佛罗里达大学）

作者：Jialu Wang, Jianing Deng, Shuqing Luo, Yuanzhe Li, Dongwei Wang, Jingtong Hu, Huanrui Yang, Song Wang, Tianlong Chen

英文摘要：Unlearning ensures LLM compliance by removing the influence of private or copyrighted training data. However, since LLM models typically undergo post-training compression, like quantization, in practical deployment, it has been observed that the unlearning effect can be substantially weakened, with the forgetting behavior degrading more severely than that of model utility. This paper proposes a quantization-robust unlearning framework that makes forgetting robust to quantization while maintaining overall model utility. We analyze this gap through the lens of loss landscape. Specifically, our analysis reveals a curvature-based criteria that pinpoints sensitive weights in the unlearned model that leads to both non-robust forgetting and reduced utility. We therefore propose sensitivity-guided noisy regularization, which is applied on the sensitive parameters to steer the model convergence towards a smoother minima of uniformly low forget and retain losses. Balancing unlearning and utility, we further propose forget-critical optimization, which updates only forget-critical layers, preserving most of the network to retain useful knowledge. Extensive experiments on the MUSE and TOFU benchmarks across multiple LLM unlearning algorithms show that our approach achieves substantially more quantization-resilient forgetting while maintaining utility.

AI 总结：针对生物多样性监测中专家标注受限的问题，综述主动学习在声学与图像模态的应用，提出兼顾训练、验证及生态推断的预算分配框架与未来路线图。

链接：https://arxiv.org/abs/2609.27409

机构：University of Amsterdam（阿姆斯特丹大学）； Tampere University（坦佩雷大学）； Leiden University（莱顿大学）； Naturalis Biodiversity Centre（自然生物多样性中心）

作者：Ben McEwen, Shiqi Zhang, Dan Stowell

英文摘要：Limited expert annotation capacity is a pervasive constraint in biodiversity monitoring. Passive acoustic recorders and camera traps generate data faster than experts can analyse them. Machine learning (ML) models can process these data at scale, but their reliability depends on the quality, quantity, and coverage of labelled samples, so expert time remains a constraint. Active learning (AL) eases this bottleneck by selecting, under a fixed annotation budget, the samples expected to improve a model most, and published evidence shows it can reduce the labels needed to reach a target performance. Monitoring programmes, however, face a broader question: how should a limited expert budget be divided so that model training, validation, and the ecological estimates built on model outputs all remain reliable? Because AL selects samples non-randomly, its labels are unsuitable for validation, calibration, or threshold selection, a tension rarely acknowledged. We synthesise AL research across acoustic and image modalities and identify gaps and opportunities. Most studies evaluate query strategies on pre-labelled benchmarks with simulated annotators; deployments in real monitoring workflows are rare and concentrate on birds and cetaceans. Bats, insects, amphibians, and fish are underrepresented, and multimodal applications remain largely unexplored. Evaluation centres on headline reductions in annotation effort, often without random-sampling baselines, per-class results, or calibration analysis, and rarely accounts for the labels required for validation. We provide a tutorial treatment of the AL loop that makes these budget decisions explicit, and a roadmap towards AL methods that support label-efficient training, validation, and trustworthy downstream ecological inference.

AI 总结：提出任务条件潜在对齐（TCLA）框架，通过共享潜在空间和任务条件分布对齐，在跨会话和跨受试者神经解码中显著提升稳定性和鲁棒性，优于现有方法。

链接：https://arxiv.org/abs/2609.27441

机构：Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）； University of Pittsburgh（匹兹堡大学）； Chinese University of Hong Kong, Shenzhen（香港中文大学（深圳））

作者：Canyang Zhao, Bolin Peng, J. Patrick Mayo, Ce Ju, Bing Liu

英文摘要：Achieving stable long-term neural decoding in invasive brain-machine interfaces (BMIs) remains challenging due to variations in recorded neural populations across sessions. Current latent alignment approaches may overlook task-dependent structure during cross-session adaptation. We propose Task-Conditioned Latent Alignment (TCLA), a framework that stabilizes neural decoding by learning a shared latent space. TCLA learns a low-dimensional source representation using neural reconstruction and continuous behavioral supervision. During target-session adaptation, the shared representation is fixed, while target neural activity is mapped into the source latent space by aligning source and target distributions separately for each task condition. We evaluated TCLA on seven nonhuman primate datasets spanning multiple tasks. In long-term cross-session evaluation, TCLA achieved a mean $R^2$ of $0.476\pm0.014$ with a negative $R^2$ failure rate of only 6.8\%. Across 1,356 within-subject session pairs, TCLA achieved a mean $R^2$ of $0.371\pm0.009$ with a failure rate of 6.8\%. Across 2,134 cross-subject session pairs, TCLA achieved a mean $R^2$ of $0.218\pm0.004$ with a failure rate of 12.9\%, substantially better than those of the comparison methods. These results demonstrate that by preserving behaviorally relevant and task-dependent latent structure, TCLA improves the robustness of neural decoding across recording sessions and subjects. The source code is publicly available at \href{ this https URL }{ this https URL }.

AI 总结：本文介绍TNLearn，一个开源Python包，用于自动构建基于任务的神经元和网络，以促进其在科研和工业中的应用，并已开源且成为PyTorch生态项目。

链接：https://arxiv.org/abs/2609.27564

机构：The City University of Hong Kong（香港城市大学）； Peking University（北京大学）； Shenzhen H&T Intelligent Control Co., Ltd.（深圳华泰智能控制有限公司）； Harbin Institute of Technology（哈尔滨工业大学）

作者：Meng Wang, Tieyun Li, Juntong Fan, Hanyu Pei, Jing-Xiao Liao, Yaodong Yang, Jianwei Ma, Fenglei Fan

英文摘要：The brain does not rely on a single type of neuron to perform all kinds of tasks; instead, it designs different neurons for different tasks. The concept of task-based neurons represents a paradigm shift compared to task-based architectures. It argues that solving a specific problem requires customized neurons, as task-based neurons capture useful prior knowledge from task-related data. To facilitate the use of task-based neurons in scientific research and industrial applications, we introduce TNLearn, an open-source Python package that provides automated construction of task-based neurons and networks, enabling smooth training of task-based networks. Comprehensive documentation, including technical exposition, API reference, and representative examples, is available online. TNLearn is open-sourced at this https URL and has become a PyTorch ecosystem project.

AI 总结：针对多模态训练中的模态不平衡问题，提出方差校准动量VCMM，通过在线估计噪声与漂移并利用卡尔曼控制器自适应调整各模态动量，在四个基准上以低开销取得一致提升。

链接：https://arxiv.org/abs/2609.27577

作者：Zhongjing Gu, Chenyang Huang, Yufa Feng, Chong He, Qinxu Ding, Yiming Cui

英文摘要： Multimodal joint training often suffers from modality imbalance, where a dominant modality suppresses the optimization of others. Existing methods mainly balance modality learning by modulating gradient magnitudes or directions, modifying optimization objectives, or adjusting training strategies, with most interventions focusing on the current update. However, when combined with widely used momentum-based optimizers, the update also incorporates accumulated information from previous gradients, which is not explicitly addressed by current-step modulation alone. To address this issue, we propose Variance-Calibrated MomentuM (VCMM), which adapts gradient memory to modality-specific gradient dynamics. Specifically, VCMM estimates minibatch noise and temporal drift online and uses their relative strength to determine modality-specific momentum through a Kalman-inspired controller. We further center the control signal across modalities and apply exact bias correction for the time-varying first moment, enabling adaptive gradient memory without extra network passes or explicit learning-rate scaling. Experiments on four multimodal benchmarks demonstrate consistent improvements with modest training overhead.

AI 总结：本文实证检验步进法则在小于59M参数的小规模语言模型上的迁移性，发现幂律形式仍成立但系数需重新校准，直接迁移会高估最优学习率约4倍。

链接：https://arxiv.org/abs/2609.27581

机构：HSE University（高等经济大学）； Faculty of Computer Science（计算机科学学院）

作者：Egor Romanyukov, Timofey Novikov, Timur Shokarov, Elizaveta Zorkina, Anastasia Palienko, Stepan Dergachev

英文摘要：Step Law gives power-law formulas for the optimal peak learning rate eta* and batch size B* when pre-training language models. It was calibrated on models between 59M and 1B parameters; the small-model regime N < 59M was never tested empirically by its authors. This regime matters for single-GPU training, interpretability research, educational experiments, and settings where larger models are infeasible on memory or cost grounds. We test whether Step Law transfers to small language models. We consider three outcomes: H1, the original coefficients work directly; H2, the power-law form holds but with different coefficients; and H3, a power law does not describe the optima in this regime. All experiments use a single nanoGPT/TinyStories pipeline with a 2048-token BPE vocabulary, AdamW, and a warmup-cosine schedule. The optimum for each (N, D) cell is extracted from the loss surface L(eta, B) via a local quadratic approximation in log-log coordinates over the smoothed training loss. The final dataset contains 29 unique (N, D) cells and 935 analysis-ready runs. The main refit uses 25 cells (815 runs) in the working range 4 <= D/N <= 600. On the pooled data we accept H2: the functional form is preserved, but the coefficients differ from the original. We obtain eta*(N, D) = 0.0985 N^(-0.508) D^(0.238) (R^2 = 0.834) and B*(D) = 3.6 x 10^(-4) D^(0.931) (R^2 = 0.950). Step Law's structural claim that B* is independent of N is reproduced (p = 0.87), but the growth of B* with D is nearly twice as steep as in the original work. Direct transfer of Step Law systematically overestimates the optimal learning rate: the median ratio eta_SL / eta* is approximately 4.0x, with a range of 2.4x to 6.6x.

AI 总结：本文提出能力流形框架，将下游能力映射到预训练、后训练和测试时资源，统一现有缩放定律，并量化能力对资源的敏感性。

链接：https://arxiv.org/abs/2609.27588

机构：University of Leeds（利兹大学）

作者：Syed Ali Raza Zaidi, Maryam Hafeez

英文摘要：Existing machine learning (ML) scaling laws relate predictive loss to compute, model parameters, and data. However, as models are increasingly deployed through agentic harnesses, loss alone is insufficient to characterize downstream performance: models with similar loss can exhibit different capabilities in reasoning, retrieval, planning, and adaptation. Yet, no unified framework connects such capabilities to the coupled resources available across the ML lifecycle. We bridge this gap by introducing a capability manifold, a multidimensional framework mapping downstream capabilities to pre-training, post-training, and test-time resources through bounded scaling functions. Analytical Jacobians quantify capability sensitivity to resource changes and interactions. As an initial application, we embed Kaplan- and Chinchilla-type scaling laws and test-time compute within the framework, demonstrating how existing scaling relationships can be unified as trajectories on a common capability manifold.

AI 总结：FLEET通过记忆机制将生成表示为熵阈值之上的稀疏轨迹，推断逐标记效用分数调整对数几率，在保持相同准确率下实现3倍加速，并将LiveCodeBench Pass@32从59.9%提升至66.2%。

链接：https://arxiv.org/abs/2609.27657

机构：Kharkiv National University of Radio Electronics（哈尔科夫国立无线电电子大学）

作者：Oleksii Streltsov, Oleksandra Vitko

英文摘要：Solutions based on large language models (LLMs) often rely on temperature sampling to improve accuracy and stability by aggregating multiple samples from the completion distribution. However, this memoryless approach is inherently suboptimal: because it lacks awareness of prior generations and their evaluations, it produces an increasing proportion of semantically duplicate answers as more samples are drawn, leading to diminishing returns. To address this limitation, we introduce FLEET, a novel method that integrates a memory mechanism into the generation process. FLEET represents each generation as a sparse trajectory through states whose entropy exceeds a predefined threshold and uses these trajectories to infer per-token utility scores that adjust the logits. Benchmark evaluations demonstrate that FLEET achieves the same accuracy as the repeated sampling baseline, with a 3x speedup, and substantially improves accuracy on complex coding tasks (LiveCodeBench Pass@32 increases from 59.9% to 66.2%) under the same budget. Furthermore, in the greedy-decoding configuration evaluated here, the approach is deterministic and uses a single calibration pass to derive its principal hyperparameters, requiring only minimal modifications to existing LLM pipelines.

AI 总结：针对私有去中心化学习中的噪声与偏差问题，提出PRDO方法，结合递归估计与精确扩散，实现非凸收敛保证并提升准确性。

链接：https://arxiv.org/abs/2609.27658

作者：Yizhao Fan, Wenjian Luo, Jiaojiao Zhang

英文摘要：Private decentralized learning is affected by sampling noise, privacy noise, and decentralized bias under heterogeneous data. We propose Private Recursive Decentralized Optimization (PRDO). PRDO uses recursive estimation with same-batch gradient differences to reduce estimation errors caused by sampling and privacy noise, while its Exact Diffusion component corrects decentralized bias arising from data heterogeneity. Our analysis establishes a nonconvex convergence bound without assuming uniformly bounded data heterogeneity across nodes. It further gives a sufficient condition under which recursive gradient differences yield strictly lower query sensitivity than private Exact Diffusion, together with an example that rigorously satisfies this condition. Experiments show improved accuracy over the evaluated baselines.

AI 总结：本文提出原位表示精炼方法，通过注意力门控更新构建任务特定预测器，RefineICL模型在多个基准上超越现有方法，并验证了支持表示更新的有效性。

链接：https://arxiv.org/abs/2609.27679

机构：Ant Group（蚂蚁集团）

作者：Tian Zhou, Beverly Jin, Linxiao Yang, Xue Wang, Wenwei Wang, Bingqing Peng, Mengni Ye, Jinjie Gu, Liang Sun

英文摘要：What reusable computation should a tabular foundation model learn when every table defines a new supervised task? We develop in-situ representation refinement: support labels guide updates to the episode's representations, and these updates transfer to unlabeled queries without changing model parameters. A regularized leave-one-out objective yields a support correction and its query extension. The leading term separates attention-based reading from state-dependent scaling, motivating RefineICL: an attention-gated, FFN-free contextual stack with selected low-rank feature interaction and typed memory. RefineICL-L24 reaches 0.93836 OVR-AUC and 0.87173 accuracy on AMLB29. A benchmark-informed continuation reaches 1644.8 Elo on the 38-dataset TabArena snapshot, 31.4 Elo above TabPFN-3 under the same evaluation. It also improves all four reported metrics over TabPFN-v3 on both TabZilla views. In a matched 100K-update depth grid, an expanded FFN gives no consistent validation benefit and uses 60.2% more peak inference memory at L8. Internal interventions show that support representations are more than a static source of labels: removing one intermediate support update, while preserving the query output, increases final query cross-entropy in all 72 tested episodes. Together, the derivation and interventions explain how attention-gated updates can construct a task-specific predictor in context.

AI 总结：本文提出牛顿-舒尔茨注意力（NS-Attn.），一种无参数变换，通过降低注意力输出矩阵的谱集中度并增加有效秩，在ViT和Swin上于CIFAR-10/100的12个匹配种子比较中平均提升0.25-0.83个百分点准确率，但增加推理延迟。

链接：https://arxiv.org/abs/2609.27735

机构：University of Exeter（埃克塞特大学）

作者：Xiaohe Jiang (1), Guoqiang Zhang (1), Tianjin Huang (1), Ronghui Mu (1) ((1) University of Exeter)

英文摘要：Newton-Schulz (NS) iteration has recently been used in the Muon optimizer to transform update matrices during the training of large language models. Motivated by its spectral effect, we investigate applying NS directly to Transformer attention representations. We introduce Newton-Schulz Attention (NS-Attn.), a parameter-free transformation applied to the output of each attention head. Each head output is arranged as a feature-by-token matrix and normalized by its Frobenius norm. We then apply a finite NS polynomial step and restore the original norm. The objective is to reduce spectral concentration and increase effective rank before standard head merging and output projection. Across ViT and Swin on CIFAR-10 and CIFAR-100, NS-Attn. improves final-epoch accuracy in all 12 matched-seed comparisons, with mean gains of 0.25--0.83 percentage points. ViT ablations show higher mean accuracy with one iteration than with two. Spectral analysis further shows reduced leading-eigenvalue concentration and increased effective rank. These gains incur additional inference latency.

AI 总结：针对异常时间序列稀缺且形态异质的问题，提出CAST框架，通过两阶段预训练与微调，利用正常数据学习动态并条件化异常结构，在多个真实数据集上优于现有方法。

链接：https://arxiv.org/abs/2609.27825

作者：Haochen Zhang, Jie Peng, Songyuan Sui, Yu-Chao Huang, Xiangqi Zhu, Tianlong Chen

英文摘要：Anomalous time series play a critical role in safety-critical domains, yet they are inherently scarce, heterogeneous, and costly to obtain. Existing time series generation methods predominantly focus on synthesizing normal data, providing limited value when anomalous samples are needed. We identify two fundamental challenges in anomaly generation: (i) the scarcity of anomaly data, and (ii) the heterogeneous morphological characteristics of anomalies. To address these challenges, we propose CAST, a Context- and Anomaly Structure-conditioned Time series anomaly generation framework with principled two-stage pretraining and finetuning strategy. In pretraining stage, we leverage abundant normal time series data to learn underlying system dynamics and substantially mitigate the limited availability of anomaly data. During finetuning, CAST explicitly conditions the generator on learned anomaly structure representations, enabling it to capture heterogeneous anomaly morphologies under similar contextual conditions. Extensive experiments on multiple real-world univariate and multivariate datasets demonstrate that CAST consistently outperforms state-of-the-art anomaly generation methods in terms of both generation fidelity and downstream task utility, highlighting the effectiveness of the proposed approach.

AI 总结：本研究证明一位无偏压缩的极小极大二阶矩下确界，揭示高斯极小极大最优性需重尾分布，并提出有限随机性逼近方案，在优化中匹配理论下界。

链接：https://arxiv.org/abs/2609.27860

机构：Key Laboratory of System Software (Chinese Academy of Sciences)（系统软件重点实验室（中国科学院））； State Key Laboratory of Computer Science（计算机科学国家重点实验室）； Institute of Software, Chinese Academy of Sciences（中国科学院软件研究所）； School of Computer Science and Technology, University of Chinese Academy of Sciences（中国科学院大学计算机科学与技术学院）

作者：Tao Jiang, Minbo Gao, Shaowei Cai

英文摘要：A pointwise-unbiased one-bit compressor reconstructs every real input in expectation while transmitting one bit. For a scalar source $P$ with CDF $F$, mean $m$, and $\mathcal J(P)=\int_{\mathbb R}\sqrt{F(r)(1-F(r))}\,dr$, we prove that the infimum of the source-averaged reconstruction second moment over all public-coin one-bit codes unbiased on $\mathbb R$ is $m^2+\mathcal J(P)^2$. For regular full-support sources, a distribution-centered random-threshold code attains this value; a converse over arbitrary randomized binary encoders and an equality analysis characterize every attaining code up to null sets, bit relabeling, and public-seed refinement. For the Gaussian location family $\mathcal N(\mu,\sigma^2)$ with $|\mu|\le c\sigma$, the equal prior on the endpoint means is least favorable and the minimax value is $\sigma^2\Lambda_c^2$. Exact Gaussian minimax optimality forces a critical heavy tail: at the endpoint means, absolute moments are finite exactly for $p<3$, and $\Pr(W>t)=\Theta(t^{-3}/\sqrt{\log t})$. A Cauchy-mixture robustification inflates the second moment by at most $1/(1-\eta)$ while making every positive-order absolute moment finite. Finite-support public randomness with finite decoder means cannot achieve exact unbiasedness on $\mathbb R$, but a bounded-output approximation using exactly $R$ shared random bits has explicit bias and second-moment bounds converging to the minimax constant. Finally, coordinate allocation communicates exactly $B$ bits per Gaussian-gradient query. On Kim's continuous quadratic hard family, the expected optimization guarantee matches the lower bound in its dependence on $(\sigma,d,B,\varepsilon)$, and a finite-variance high-probability bound incurs only a logarithmic confidence factor.

AI 总结：本研究提出双轴漂移检测门，结合功能轴与新颖性轴，解决深度交换分数混淆问题，在广义类别发现中实现高AUROC，并验证于多种骨干网络。

链接：https://arxiv.org/abs/2609.27865

机构：Kagoshima University（鹿儿岛大学）

作者：Kentaro Oda

英文摘要：Sharing a deep encoder does not, by itself, fix the central confound of task-comparison scores. We show that cross-evaluated heads on a frozen shared representation inherit the extrapolation confound of shallow exchange scores: pure input rotations with fixed labels inflate a deep exchange score from about 0 to 0.80, while representation-novelty scores are blind in the complementary direction (flat under label permutations that change the task completely). Transplanting a conditional two-discriminator discrepancy into the embedding space resolves both blind spots: the functional axis stays within +-0.001 under rotations and tracks label-permutation drift mass monotonically. Built into a mixture-of-heads lifecycle, the two-axis gate attains better decision quality with fewer heads than exchange or novelty triggers at a matched training budget. On generalized category discovery, the same chunk-level functional axis separates semantic novelty from photometric shift with AUROC 0.98-0.99 where per-input OOD scores (MSP, Energy, Mahalanobis, KNN) sit near chance for that distinction. All findings replicate across frozen ImageNet-21k ViT-B/16 and self-supervised DINOv2 backbones on CIFAR-100, and extend to residual adapter pools with recurrence, where a null-calibrated novelty trigger never fires on mechanism changes while the two-axis gate handles them with full recurrence reuse. We state explicitly the common-factoring condition under which embedding-space conclusions transfer to the original mechanism.

AI 总结：本研究提出条件双判别器嵌入差异，解决深度专家池中共享编码器的任务比较混淆，以双轴门实现更优决策质量，并在广义类别发现中达到高AUROC。

链接：https://arxiv.org/abs/2609.27866

机构：Kagoshima University（鹿儿岛大学）

作者：Kentaro Oda

英文摘要：Sharing a deep encoder does not, by itself, fix the central confound of task-comparison scores. We show that cross-evaluated heads on a frozen shared representation inherit the extrapolation confound of shallow exchange scores: pure input rotations with fixed labels inflate a deep exchange score from about 0 to 0.80, while representation-novelty scores are blind in the complementary direction (flat under label permutations that change the task completely). Transplanting a conditional two-discriminator discrepancy into the embedding space resolves both blind spots: the functional axis stays within +-0.001 under rotations and tracks label-permutation drift mass monotonically. Built into a mixture-of-heads lifecycle, the two-axis gate attains better decision quality with fewer heads than exchange or novelty triggers at a matched training budget. On generalized category discovery, the same chunk-level functional axis separates semantic novelty from photometric shift with AUROC 0.98-0.99 where per-input OOD scores (MSP, Energy, Mahalanobis, KNN) sit near chance for that distinction. All findings replicate across frozen ImageNet-21k ViT-B/16 and self-supervised DINOv2 backbones on CIFAR-100, and extend to residual adapter pools with recurrence, where a null-calibrated novelty trigger never fires on mechanism changes while the two-axis gate handles them with full recurrence reuse. We state explicitly the common-factoring condition under which embedding-space conclusions transfer to the original mechanism.

AI 总结：本研究揭示自主科学发现中因错误配对导致的虚假科学归纳现象，提出错误连贯性为预算错配主因，并引入隔离与分流机制以保障实验有效性。

链接：https://arxiv.org/abs/2609.27883

机构：Changchun University of Science and Technology（长春理工大学）

作者：Hanbing Liang, Fujun Liu

英文摘要： Closed-loop discovery systems increasingly execute experiments and update decisions autonomously, turning record integrity into part of the experimental apparatus. We show that false-science induction arises when legitimate physical objects and measurements are paired incorrectly, driving neural surrogates to faithfully learn record-induced associations that do not correspond to the true object-outcome relationship while marginal data distributions remain unchanged. Across green fluorescent protein fitness and materials band-gap prediction loops, coherent paired misbinding systematically redirects experimental budgets toward low-performing basins, whereas same-volume random swaps have negligible effects. These observations identify error coherence, rather than raw error frequency, as the primary variable controlling this budget misallocation in the tested loops. The resulting binding identifiability boundary supports monitored-axis quarantines and feedback-conflict triage, which intercept over-concentrated proposals before execution and isolate the corrupted hypothesis axis.

AI 总结：本研究揭示全局树模型在预测层级聚合值时因超出训练范围而崩溃，低估总量达30-496倍，并提出按序列缩放、加权聚合行和季节性差分等预防方法及三步诊断程序。

链接：https://arxiv.org/abs/2609.27912

机构：Field Nation LLC（Field Nation 有限责任公司）

作者：Md Rezwanul Islam, Wael Mohammed

英文摘要：Global forecasting models pool many series and learn one shared function. Gradient-boosted trees are their most common form. We measure a failure of this design that has not, to our knowledge, been documented. Train a global tree on the individual series of a hierarchy, then ask it for the hierarchical aggregate. The aggregate sits far outside the model's training range, and the forecast collapses. The model under-predicts the total by 30-50x in our production deployment, and by up to 496x in a public M5 reconstruction. The mechanism is known: beyond its training range, a tree predicts a constant. It surfaces at the aggregate because the total dwarfs every training series. The cure is not new. Per-series scaling, the preprocessing step that Montero-Manso and Hyndman (2021) recommend, prevents the collapse. So do a weighted aggregate-level training row and seasonal differencing. Our contribution is the characterization. The collapse reproduces on five panels: a production business-to-business marketplace, a synthetic hierarchy, M5, Australian Tourism, and a public business-buyer panel. It holds on three tree libraries, is invariant across training seeds, and is statistically significant. Its onset is immediate and tracks a simple support bound: a scale gap of only 1.15x already costs a third of the total. No standard configuration change prevents it: pooling every hierarchy level into training fails at scale, and the one knob that fits linear models in the leaves softens it without curing it. Rolling the forecasts forward recursively separates the cures: the aggregate-row cure re-collapses, per-series scaling degrades but stays low, and only seasonal differencing keeps its one-step accuracy unchanged. We close with a three-step procedure for diagnosing and preventing the failure in deployed systems.

AI 总结：针对专家可靠性随输入变化且易冲突的场景，提出基于概率电路的动态融合框架，利用上下文可信度聚合黑盒专家意见，无需重训练，在多项选择任务中优于静态基线。

链接：https://arxiv.org/abs/2609.27913

机构：The University of Texas at Dallas（德克萨斯大学达拉斯分校）； TU Darmstadt（达姆施塔特工业大学）； The University of Texas at Tyler（德克萨斯大学泰勒分校）； Air Force Research Lab（空军研究实验室）

作者：Pranuthi Tenali, Sahil Sidheekh, Saurabh Mathur, Vijayalakshmi Saravanan, Erik Blasch, Kristian Kersting, Sriraam Natarajan

英文摘要：We study the problem of aggregating opinions from multiple black-box experts in noisy, conflict-prone settings where expert reliability varies across inputs. Static aggregation methods, such as majority voting, fail to capture this variability and often yield unreliable outcomes under disagreement. We propose a tractable, probabilistic-circuit-based fusion framework that dynamically combines expert responses using context-specific credibility estimates, enabling principled and reliable reasoning. The framework is agnostic to the underlying experts and does not require access to their internal representations or any retraining. We empirically validate our approach on multiple-choice question answering tasks using multiple LLMs as experts, comparing against individual models and static ensemble baselines. Our method consistently improves predictive performance and produces more reliable decisions under conflict, highlighting the effectiveness of context-aware credibility modeling for robust multi-expert fusion.

AI 总结：本文通过预注册实验，在旅行商问题上研究测试时预算重新分配的决定因素，发现实例难度差异而非平均难度是关键，且预算感知策略可恢复大部分改进。

链接：https://arxiv.org/abs/2609.27917

机构：Hankuk University of Foreign Studies（韩国外国语大学）

作者：Jinhyung Bae

英文摘要：Neural combinatorial optimization solvers generate many candidate solutions per instance and report the best one found, using the same sample budget for every instance regardless of difficulty. A companion study showed that reallocating a fixed budget toward harder instances can improve solution quality, but that the standard way of measuring this improvement is biased: deciding an allocation and evaluating it on the same data can manufacture an apparent gain even when none exists. This left open what property of a workload determines whether reallocation is worth doing, and whether a policy that spends part of the budget to decide how to allocate the rest still pays once that cost is counted. This paper answers both questions through pre-registered confirmatory experiments -- analysis and verdict criteria fixed before data collection -- across three independently trained solvers and two ways of constructing harder workloads on the traveling salesman problem. Within the workloads we study, the deciding property is how varied the instances within a workload are in difficulty, not how difficult the workload is on average: a uniformly easy or uniformly hard workload offers little room for reallocation, while a mixed workload offers substantial room. A budget-aware policy that pays for its own information about instance difficulty recovers most, though not all, of the improvement available when that information is assumed free. Every experiment was independently recomputed from its written specification, and every correction to an earlier version -- including two that weakened the paper's own claims -- is reported with the direction it moved the conclusion. The paper offers a specific empirical answer and a template for verifying that answer is not an artifact of how it was measured.

AI 总结：本文证明二值量化神经网络训练（2-QNNT）以输入和输出维度之和为参数时是 W[1]-困难的，即使在零误差和特定前缀链数据上也成立，并基于指数时间假说排除了 FPT 算法，核心归约利用单翻转路由等价性从边不相交路径问题构造。\n

链接：https://arxiv.org/abs/2609.27932

机构：Key Laboratory of System Software (Chinese Academy of Sciences)（系统软件重点实验室（中国科学院））； State Key Laboratory of Computer Science（计算机科学国家重点实验室）； Institute of Software, Chinese Academy of Sciences（中国科学院软件研究所）； School of Computer Science and Technology, University of Chinese Academy of Sciences（中国科学院大学计算机科学与技术学院）

作者：Tao Jiang, Minbo Gao, Shaowei Cai

英文摘要：Ganian et al. (ICLR 2026) proved that quantized neural network training is fixed-parameter tractable when parameterized jointly by architecture treewidth, input dimension $\alpha$, and output dimension $\omega$, and left open whether $\alpha+\omega$ alone yields fixed-parameter tractability. We prove that 2-QNNT is W[1]-hard parameterized by $\alpha+\omega$. The hardness already holds with zero error on $D_k=\{(\xi^{(r)},\xi^{(r)}):0\le r\le k\}$, where every input equals its target, $|D_k|=\alpha=\omega=k+1$, and the examples form a coordinatewise prefix chain. It also holds when every non-source bias is fixed to zero. Under the Exponential Time Hypothesis, no algorithm runs in $f(\alpha+\omega)|I|^{o(\alpha+\omega)}$ for any computable $f$. The reduction starts from DAG edge-disjoint paths, converts edge capacity to vertex capacity with a directed line graph, and normalizes the result into a valid layered architecture. The key structural step is a one-flip routing equivalence: on the prefix-chain inputs, nonnegative binary weights make every activation monotone, and each required output transition has a weight-one predecessor making the same transition. Iterating this relation backward extracts a path from the unique changing input, while different transitions yield vertex-disjoint paths. In particular, every neuron on these inputs has only $k+1$ possible activation profiles.

AI 总结：针对LLM裁判预测集在流量转移下覆盖不足的问题，提出CS-WCP方法，通过构建源目标组质量精确区间并取加权共形集并集，在不确定组比例下提供可审计的鲁棒覆盖保障，实验验证其高覆盖率与保守尾部保护。

链接：https://arxiv.org/abs/2609.27955

机构：Iowa State University（爱荷华州立大学）； Kalinga Institute of Industrial Technology（卡林加工业技术学院）

作者：Ibne Farabi Shihab, Fariya Afrin

英文摘要：Prediction sets built from an LLM judge can undercover when deployment traffic changes the prevalence of task or policy groups. Weighted conformal prediction is exact under covariate shift when the density ratio is known, but group proportions must usually be estimated from finite unlabeled samples. We introduce confidence-set weighted conformal prediction (CS-WCP), which constructs simultaneous exact intervals for source and target group masses and returns the union of weighted conformal sets over every compatible ratio vector. For a fixed or independently learned finite partition, CS-WCP attains coverage at least 1-alpha-delta_w-tau_A-kappa, where tau_A measures within-cell covariate mismatch and kappa measures conditional shift. A linear endpoint rule computes the robust union in O(G|Y|) time. Across 336 constructed shared-support traffic shifts, CS-WCP reaches 0.973 mean coverage with 13 point failures, compared with 0.954 and 44 failures for source conformal prediction, at mean binary set sizes 1.74 and 1.65. On 336 natural cross-task transfers, coverage rises from 0.882 to 0.962, but mean set size reaches 1.87 and a size-matched group plug-in baseline is competitive. The method therefore supplies an auditable coverage safeguard under uncertain mixture weights; its value is conservative tail protection, not scalar probability calibration or uniformly smaller sets.

AI 总结：提出I-SplineFlow，用积分单调样条参数化随机插值调度器，实现扩散与流模型的少步生成，在低NFE下改善FID，且训练仅需数分钟。

链接：https://arxiv.org/abs/2609.27963

作者：Md Sakib Hossain Shovon, Md Rifat Ur Rahman, Md Abtahi Majeed Chowdhury, Yunhong Min, Jaesik Choi, Minhyuk Sung

英文摘要：Few-step generation with pretrained diffusion and flow models can be accelerated by lightweight training that optimizes the sampling trajectory rather than the network. A recent approach parameterizes the stochastic interpolant (SI) scheduler as a smooth curve whose control points enforce the three properties an SI scheduler must satisfy: fixed boundary conditions, a monotone signal-to-noise ratio (SNR), and differentiability. Existing parameterizations use globally supported polynomial bases, where every control point moves the whole curve and higher expressiveness needs a higher degree, which couples distant regions of the schedule during optimization. We introduce \emph{I-SplineFlow}, which parameterizes the scheduler with integrated monotone splines (I-splines). I-splines decouple the polynomial degree from the number of mixture weights, so support width and smoothness can be chosen per model at a fixed weight count, and the compactly supported derivative basis makes the scheduler Jacobian orders of magnitude better conditioned than a Bézier basis. Boundary conditions and a strictly monotone SNR hold by construction, with no ordering constraint on the parameters and closed-form velocity derivatives. Across diffusion (EDM) and flow (ReFlow, Simple ReFlow) models, I-SplineFlow improves few-step FID over Bézier scheduling in most settings, most clearly at the lowest NFEs, and trains in minutes. Ablations show that both the degree freedom and the monotonicity constraint are needed. The code will be released upon acceptance.

AI 总结：针对电池剩余放电时间预测中负载未知的难题，提出相对放电阶段（RDS）分类指标，结合SOC估计与轻量级时序网络，在两个公开数据集上实现超过80%的分类准确率。

链接：https://arxiv.org/abs/2609.27986

机构：Ton Duc Thang University（孙德胜大学）； AIWARE Limited Company（AIWARE有限公司）； Van Lang University（万朗大学）

作者：Khoa Tran, Tri Le, Hung-Cuong Trinh, Hung Tran-Nam

英文摘要： Accurate remaining discharge time (RDT) prediction is challenging in real-world battery applications because future load profiles are unknown and highly dynamic. To address the uncertainty of continuous RDT regression, this paper introduces Relative Discharge Stage (RDS), a battery-management indicator that represents the remaining discharge condition using five interpretable classes: Normal, Good, Moderate, Low, and Recharge Required. Unlike state of charge (SOC), which reflects the current charge level, RDS characterizes the remaining discharge process without requiring future-current information during inference. A physics-informed RDS classification framework is proposed, combining SOC estimation with lightweight temporal learning. The SOC-estimation component includes second-order ECM state and terminal-voltage prediction, hysteresis and OCV temperature correction, core-temperature estimation, and AEKF state correction, supported by OCV evaluation, online STC-ECM parameter adaptation, and pretrained neural residual-voltage correction. The measured current, terminal voltage, surface temperature, and estimated SOC are arranged into a sliding observation window and processed by a lightweight temporal convolutional network. Experiments on two public lithium-ion battery datasets demonstrate robust RDS classification, with accuracy exceeding 80% under varying load and thermal conditions.

AI 总结：针对中小语言模型工具智能体在长时程任务中的结构性错误，提出基于异构图记忆的失败感知检索框架FRESH，通过结构化经验提升任务成功率与工具使用可靠性。

链接：https://arxiv.org/abs/2609.28003

机构：Southeast University（东南大学）

作者：Jiaxing Li, Lei Song, Rui Dong, Youyong Kong

英文摘要：Small and medium-sized language models offer cost-effective executors for tool-using agents, making them attractive for local and large-scale deployment. However, in long-horizon and stateful environments, they often make structural errors such as missing required observations, performing premature writes, repeating failed calls, and violating action preconditions. These errors can lead to incorrect state updates, policy violations, and costly or irreversible consequences, making reliable tool execution a critical deployment challenge. Existing fine-tuning approaches require substantial data and computation, while flat memory may retrieve failed actions without preserving their causal context or safety conditions. In this paper, we propose FRESH, a Failure-aware Retrieval framework over Experience-Structured Heterogeneous graphs, which transforms historical successes and failures into structured external experience for tool-using agents. By explicitly modeling the dependencies among tasks, actions, errors, repairs, and execution conditions, FRESH helps frozen language models reuse reliable strategies, avoid recurring failures, and make safer decisions in stateful tool interactions. Experiments on $\tau$-Bench and AppWorld with multiple open-source models show that FRESH consistently improves task success and tool-use reliability over no-memory agents and representative memory-based baselines.

AI 总结：本研究探讨在仅解码器Transformer中，共享全局KV缓存时保留层特定局部历史的价值，发现局部历史能降低困惑度，并提出了减少上层构建工作的后缀调度方法。

链接：https://arxiv.org/abs/2609.28006

作者：Xinglang Xian

英文摘要：Decoder-only Transformer language models cache keys and values (KV) to reuse past computation during generation. Sharing KV across layers saves storage but reduces the diversity of representations available across depth. We study what local memory should retain alongside shared global KV, separating historical content from the input source used to form it. At 126M parameters and 2K context, an eight-seed study finds about 1.4% lower held-out test perplexity with local history than with a current-token local branch. Capacity, entry-count and training-compute controls support the value of historical content. In a two-seed comparison, this value persists when adjacent layers share local inputs while retaining independent projections; source sharing also shortens exact cache-construction dependencies. Against GQA and adjacent-layer KV sharing, equal bounded learning-rate searches and new-seed confirmation yield better same-source likelihood with larger caches and higher long-request latency. The ordering against adjacent-layer sharing persists after equal-token adaptation to 8K, with a short-context cost. The eight-seed external-book history effect remains uncertain, and downstream outcomes vary by task. We derive a sufficient suffix schedule that reduces upper-layer construction work while preserving the complete cache in exact arithmetic.

AI 总结：LAYERSCOPE提出无标签逐层几何度量框架，用于比较视频和多模态模型表示，发现中间层优于最终层，且RankMe对分类聚类最有效。

链接：https://arxiv.org/abs/2609.28086

机构：University of Melbourne（墨尔本大学）； Johns Hopkins Applied Physics Lab（约翰斯·霍普金斯大学应用物理实验室）； Johns Hopkins University（约翰斯·霍普金斯大学）； Human Language Technology Center of Excellence（人类语言技术卓越中心）； Monash University（莫纳什大学）

作者：Sandra Arcos-Holzinger, Debashish Chakraborty, Rohita Mocharla, Will Walden, Andrew Yates, Reno Kriz, Sarah M. Erfani, James Bailey, Vishal M. Patel, Sanjeev Khudanpur

英文摘要：We propose LAYERSCOPE, a label-free, layerwise framework that aims to characterize a model's learned representations in video and multimodal settings. Evaluating downstream performance using representations from final or intermediate layers typically requires large amounts of labeled data, repeated task-specific evaluations, and substantial computation. To address these limitations, LAYERSCOPE uses local, global, distributional, and correspondence-based geometric metrics to compare layerwise representation structure within and across models without requiring task-specific labels. We evaluate seven architecturally diverse models across video and multimodal classification, clustering, and text-to-video retrieval tasks from MVEB/MVEB+. We find that intermediate-layer representations can outperform final-layer and model-default outputs. We also find that no single geometric metric consistently predicts downstream performance, but note that distinct layerwise geometric signatures emerge across model families. LID shows task-dependent relationships with performance, while RankMe provides the strongest measure for classification and clustering, but is not a universal layer selector. We also find that pairing-aware metrics explain retrieval better than distributional distances alone. LAYERSCOPE therefore offers a framework for comparing representations across models and layers, enabling a more systematic evaluation in video and multimodal settings.

AI 总结：针对多模态优化中置信度不对称问题，提出MaxCR正则化方法，通过动态干预模态语义置信度，提升多模态分类整体性能。

链接：https://arxiv.org/abs/2609.28165

机构：Nanjing University of Science and Technology（南京理工大学）； Alibaba Group（阿里巴巴集团）

作者：Longfei Huang, Xiangyu Wu, Yang Yang

英文摘要：Multimodal learning (MML) falls into the optimization dilemma due to the modality imbalance phenomenon, leading to suboptimal overall performance in practice. While many attempts primarily focus on balancing the optimization dynamics across modalities to address this issue, we identify a subtle yet critical flaw: optimization yields asymmetric gains in predictive certainty, with the strong modality more confident than the weak one, driving imbalanced modality contributions. In this paper, our analysis reveals that this flaw stems from unimodal characteristics rather than multimodal learning, and this confidence discrepancy can be corrected by positive cross-modal intervention. Based on this insight, we propose multimodal Max Confidence Regularization (MaxCR) to dynamically intervene in modality semantic confidence. Specifically, the semantic confidence of each modality is tracked using a nonlinear sparsity measure. We then design max suppression and max excitation based on this measure to regularize strong and weak modalities, respectively. They penalize and encourage the top-1 confidence, thereby constraining multimodal prediction. To this end, strong and weak modalities are expected to make calibrated confidence, thereby improving the overall performance. Empirical experiments on widely used datasets reveal the superiority of our method through comparison with various state-of-the-art (SOTA) multimodal learning baselines.

AI 总结：本研究利用地理空间基础模型嵌入和Sentinel特征在罗马尼亚南喀尔巴阡山脉检测原始森林，发现缓冲空间验证削弱GFM优势，TESSERA在无缓冲下最优，但缓冲后优势消失，强调空间验证对模型迁移的重要性。

链接：https://arxiv.org/abs/2609.28194

机构：University of Cambridge（剑桥大学）

作者：Thomas Ratsakatika (1), Mihai Zotta (2), Srinivasan Keshav (3), Emily R. Lines (1) ((1) Department of Geography, University of Cambridge, Cambridge, UK, (2) Fundatia Conservation Carpathia, Brasov, Romania, (3) Department of Computer Science and Technology, University of Cambridge, Cambridge, UK)

英文摘要：Old-growth forests develop over centuries under minimal anthropogenic disturbance, producing structurally complex and biodiverse stands. In Europe, protecting them requires mapping that is accurate for individual forest parcels yet deployable continent-wide. Geospatial foundation model (GFM) embeddings enable label-scarce land classification, but their value for old-growth detection remains unknown. Here, we map old-growth forests across 211,893 ha of Romania's Southern Carpathians, a beech-spruce landscape typical of the Alpine Biogeographic Region. We construct high-confidence, expert-informed reference labels for old-growth and non-old-growth parcels. We add AlphaEarth, TESSERA v2 and Sentinel-1/2 features to a common baseline of topographic and human-access predictors, then compare them under spatially blocked validation with and without 10 km train-test buffers to limit residual autocorrelation. With buffering, GFM and Sentinel-1/2 predictors increase precision-recall AUC by 0.21-0.25 [95% CIs: 0.15-0.34] relative to baseline, indicating spectral data contain a spatially robust old-growth signal. With a PR-AUC of 0.84 [0.79-0.88], TESSERA outperforms Sentinel-1/2 (+0.08 [+0.05 to +0.11]) and AlphaEarth (+0.08 [+0.04 to +0.12]) under unbuffered spatial validation. At a 10 km buffer, however, this advantage narrows to +0.04 [-0.01 to +0.11] and +0.03 [-0.04 to +0.10], intervals consistent with no difference. At 10 m resolution, convolutional neural networks add no benefit over pixel-based XGBoost. Comparisons with four national- and continental-scale products show the importance of non-old-growth labels, and reveal 81% agreement between our predictions and a field-calibrated map. We conclude that buffered spatial validation is vital when transferring old-growth detection models to unseen landscapes, and provide our labels and predictions for future work.

AI 总结：本研究提出将平衡树递归算子扩展至自回归预测，实现对数深度和线性运行时间的语言建模，实验表明其长度外推稳健且性能接近ALiBi Transformer。

链接：https://arxiv.org/abs/2609.28212

机构：Imperial College London（帝国理工学院）

作者：Yiqin Wang, Nuri Cingillioglu, Charles Pert

英文摘要：Language modeling using Transformers has become commonplace despite their fixed computational depth and quadratic runtime with respect to input tokens. Recurrent models on the other hand offer linear depth but no parallel execution. In this work, we extend balanced-tree recursive operators from sequence encoding to autoregressive prediction, enabling all prefix representations to be computed with logarithmic depth and linear runtime. Our experiments provide an initial characterization of this model class, demonstrating robust length extrapolation and performance approaching that of ALiBi-based Transformers, highlighting its potential as an alternative architecture for language modeling.

AI 总结：hyperbolix 是 JAX 中首个全面的双曲深度学习库，提供多种流形、层族和优化器，并通过无取消公式在远距离保持 float32 精度。

链接：https://arxiv.org/abs/2609.28248

机构：University of Vienna（维也纳大学）

作者：Timo Klein, Thomas Lang, Yllka Velaj, Sebastian Tschiatschek

英文摘要： We present hyperbolix, an open-source library for hyperbolic deep learning in JAX, built on Flax NNX. To our knowledge, it is the first comprehensive, general-purpose hyperbolic deep learning library in JAX. It includes six manifolds with a common interface: Euclidean space, the Poincaré ball, the hyperboloid, the $\kappa$-stereographic model, mixed-curvature product spaces, and the proper velocity space. We implement layer families that cover linear layers, convolutions, attention, normalization, positional encoding, regression, and vector quantization. These building blocks span methods ranging from Ganea's original hyperbolic neural networks to recent fully hyperbolic architectures such as Hypformer and Lorentzian ResNet. Additionally, hyperbolix contains Riemannian optimizers implemented as optax transformations, wrapped distributions, and hyperbolic dimensionality-reduction techniques. Its API uses idiomatic JAX: Manifolds are stateless, with curvature being passed at call time, while manifold operations act on single points, with this http URL enabling batch operations. The precision of every checked operation is tested against a closed-form NumPy/SciPy transcription from the source paper or a finite difference, for both float32 and float64. On the hyperboloid, standard formulas for two-point operations, such as the distance, lose precision far from the origin, because they subtract two large, nearly equal terms. hyperbolix replaces these subtractions with cancellation-free formulas that stay accurate in float32 at distances where prior implementations return NaN. hyperbolix is available under the MIT license at this https URL.

AI 总结：提出记忆注意力（MA）方法，用令牌索引记忆替代值投影并结合上下文键构建注意力值，在匹配训练预算下提升语言建模与下游性能。

链接：https://arxiv.org/abs/2609.28399

作者：Jiale Kang

英文摘要：Language models typically construct attention values from contextual hidden states, even when some of their content may be reusable across contexts. We investigate whether token-indexed memory can replace the dedicated value projection when complemented by contextual information. We propose Memory Attention (MA), which forms values by combining layer-specific token memory with contextual keys. The memory supplies token-specific representations, while the keys preserve context dependence. At inference, normalization can be folded into the memory tables, reducing value construction to lookup and addition. Token-indexed retrieval also enables CPU offloading with prefetching, reducing GPU parameter storage. Under matched training token budgets and with additional memory parameters, experiments across attention configurations show improved language modeling and average downstream performance.

AI 总结：本文提出可微高斯动力学（DGD），通过高斯混合表示、可微聚合和反馈循环，从聚合观测中学习集体响应分布，在多个数据集上显著降低预测误差，验证了学习群体表示及反馈过程的价值。

链接：https://arxiv.org/abs/2609.28405

机构：OranAI（奥兰AI）； OranAI Ltd.（奥兰AI有限公司）； Northeastern University（东北大学）

作者：Jianxiang Ma, Mingfu Zhang, Xiaocui Yang, Yichen Gao, Junzhao Huang, Yuesong Hou

英文摘要：Collective responses depend on individual differences, contact opportunities, and accumulated experience. Learning their dynamics from aggregate counts requires connecting a population's response distribution to both current observations and future behavior. We introduce Differentiable Gaussian Dynamics (DGD), which learns this connection through three components: a Gaussian mixture representing heterogeneous response propensities, differentiable aggregation of contact intensity and behavioral probabilities, and feedback recurrence that updates subsequent responses. Reparameterized integration and temporal recurrence let aggregate prediction errors jointly train the distribution, observation functions, and feedback parameters. On four windows from KuaiRand-Pure and Online Retail II, DGD achieves lower joint behavioral negative log-likelihood than a DeepAR adaptation with a joint-behavior head. In Retail 2010, its one-day behavioral-count MAE is 4.71 versus 6.88 for this adaptation. Learning the distribution reduces behavioral negative log-likelihood by 10.82% relative to a fixed Gaussian in KuaiRand's standard-recommendation window; removing feedback dynamics raises joint KL from 0.0340 to 0.2577 in a controlled experiment. These results establish the value of learning population representations and their feedback process from aggregate observations. Code is available at this https URL.

AI 总结：本文研究单变量两层ReLU网络的最小范数分类，刻画了有无偏置惩罚时的精确解，并证明添加跳跃连接使所有KKT点全局最优，同时揭示了稀疏性限制。

链接：https://arxiv.org/abs/2609.28438

机构：University of Warsaw（华沙大学）； University of Warwick（华威大学）； INRIA & LMO, Université Paris-Saclay（法国国家信息与自动化研究所与巴黎萨克雷大学LMO）

作者：Karolina Drabik, Ben Lewis, Antoni Puch, Etienne Boursier, Piotr Hofman, Matthias Englert, Ranko Lazić

英文摘要：We study minimal-norm interpolation and $\ell_2$-regularized logistic-loss minimization for binary classification by univariate two-layer ReLU networks. We give complete geometric characterizations of the optimal classifiers in function space, resolving how the solutions depend on whether hidden-layer biases are included in the parameter norm. When biases are unpenalized, the minimal-norm interpolators are exactly the continuous piecewise-affine functions that hug every label switch and have kinks of the appropriate convexity. When biases are penalized, the minimizer is unique in function space, has exactly one kink in each intermediate same-label segment, and is therefore a sparsest positive-margin classifier. We further show that adding a free affine skip connection leaves these function-space solutions unchanged but fundamentally improves the parameter-space landscape: every KKT point of the constrained problem becomes globally optimal, whereas suboptimal KKT points can occur without the skip connection. We establish analogous global-optimality and geometric results for sufficiently weak $\ell_2$-regularization of the logistic loss. In the unpenalized-bias case, we identify an additional sparsity-like restriction, implying that most minimal-norm interpolators cannot arise as small-regularization limits of margin-normalized logistic-loss minimizers. Numerical experiments across varying dataset complexity and network width support the predicted landscape and sparsity phenomena.

AI 总结：研究数学推理中答案不变性与表示不变性的区别，发现模型对重排规则顺序的表示区分度与准确性正相关，Spearman相关达0.86。

链接：https://arxiv.org/abs/2609.28442

机构：Princeton University（普林斯顿大学）

作者：Zhixu Silvia Tao

英文摘要： Reordering a set of mathematical rules without changing its meaning should preserve the correct answer, but must a model's internal representations stay invariant too? We investigate this question using synthetic multi-step function-composition problems, each presented under multiple rule orderings with the same correct answer. We measure accuracy and permutation signal-to-noise ratio (SNR), which quantifies how distinctly ordering patterns are represented relative to variation across problem instances. Across 16 language models ranging from 1B to 8B parameters, we find a pattern: models that solve reordered problems more accurately represent different rule orderings more distinctly. Layer-averaged permutation SNR is positively rank-correlated with accuracy in every synthetic setting we evaluate, with Spearman correlations reaching 0.86. These findings highlight a distinction between answer invariance and representation invariance: successful mathematical rule composition can accompany distinct internal representations between equivalent rule orderings. This motivates distinguishing answer invariance from representation invariance, and offers a representational perspective on mathematical reasoning beyond answer accuracy alone.

AI 总结：本文提出更紧的转导局部复杂度（STLC）方法，通过 Bernstein 型不等式和剥离论证获得与归纳 Rademacher 复杂度相同的超额风险界，在可实现学习和核学习中匹配或改进现有转导结果。

链接：https://arxiv.org/abs/2609.28459

机构：Arizona State University（亚利桑那州立大学）

作者：Yingzhen Yang

英文摘要：We introduce Sharper Transductive Local Complexity (STLC), a localized complexity method for transductive learning under uniform sampling without replacement. The construction starts from a Bernstein-type concentration inequality for the supremum of the test--train empirical process. Its proof uses the modified log-Sobolev inequality for the swap walk and a two-parameter entropy closure. A peeling argument with a surrogate localization functional then gives excess-risk bounds with the same fixed-point and confidence terms as the classical inductive local Rademacher-complexity bounds, without the additional logarithmic confidence factor in earlier transductive results. For realizable learning over a binary class of VC dimension $\dVC$, with training size $m$, test size $u$, and $u\ge m\ge\dVC$, STLC yields $\cO\{\dVC\log(me/\dVC)/m\}$. This matches the standard inductive rate and, when $m\ge9$, is within a logarithmic factor of the transductive minimax lower bound of order $\dVC/m$. For transductive kernel learning, STLC gives a spectrum-adaptive excess-risk bound without the multiplicative imbalance factors appearing in the earlier local-complexity bound.