|
创作新主题 |
社区所有版块导航
DATA
| docker Elasticsearch |
WEB开发
| linux MongoDB Redis DATABASE NGINX 其他Web框架 web工具 zookeeper tornado NoSql Bootstrap js peewee Git bottle IE MQ Jquery |
机器学习
| 机器学习算法 |
产品
| 短视频 |
印度
| 印度 |
一周十大热门主题
点击阅读原文访问arxivdaily.com，涵盖CS|物理|数学|经济|统计|金融|生物|电气领域，更有搜索、收藏等功能！
cs.LG 方向，今日共计121篇
大模型相关(11篇)
【1】Confidence-Credibility Aware Weighted Ensembles of Small LLMs Outperform Large LLMs in Emotion Detection
标题：信任度-可信度感知小型LLM的加权集合在情绪检测中优于大型LLM
链接：https://arxiv.org/abs/2512.17630
备注：Accepted at IRICT 2025
摘要：本文介绍了一个基于信任加权，可信度感知集成框架的文本情感检测，孔多塞的陪审团定理（CJT）的启发。与通常依赖于同构架构的传统集成不同，我们的方法结合了架构上不同的基于小型转换器的大型语言模型（sLLM）- BERT，RoBERTa，DistilBERT，DeBERTa和ELECTRA，每个模型都针对情感分类进行了完全微调。为了保持误差的多样性，我们最小化参数收敛，同时利用每个模型的独特偏差。双加权投票机制集成了全局可信度（验证F1得分）和局部置信度（实例级概率），以动态加权模型贡献。在DAIR-AI数据集上的实验表明，我们的可信度-置信度集成达到了93.5%的宏观F1分数，超过了最先进的基准，并且显著优于大型LLM，包括Falcon，Mistral，Qwen和Phi，即使在特定任务的低秩适应（LoRA）之后。总共只有5.95亿个参数，我们的小型LLM集成证明比高达7 B参数的模型更具参数效率和鲁棒性，这表明精心设计的小型微调模型集成可以在专门的自然语言处理（NLP）任务（如情感检测）中优于更大的LLM。
摘要：This paper introduces a confidence-weighted, credibility-aware ensemble framework for text-based emotion detection, inspired by Condorcet's Jury Theorem (CJT). Unlike conventional ensembles that often rely on homogeneous architectures, our approach combines architecturally diverse small transformer-based large language models (sLLMs) - BERT, RoBERTa, DistilBERT, DeBERTa, and ELECTRA, each fully fine-tuned for emotion classification. To preserve error diversity, we minimize parameter convergence while taking advantage of the unique biases of each model. A dual-weighted voting mechanism integrates both global credibility (validation F1 score) and local confidence (instance-level probability) to dynamically weight model contributions. Experiments on the DAIR-AI dataset demonstrate that our credibility-confidence ensemble achieves a macro F1 score of 93.5 percent, surpassing state-of-the-art benchmarks and significantly outperforming large-scale LLMs, including Falcon, Mistral, Qwen, and Phi, even after task-specific Low-Rank Adaptation (LoRA). With only 595M parameters in total, our small LLMs ensemble proves more parameter-efficient and robust than models up to 7B parameters, establishing that carefully designed ensembles of small, fine-tuned models can outperform much larger LLMs in specialized natural language processing (NLP) tasks such as emotion detection.
【2】GreedySnake: Accelerating SSD-Offloaded LLM Training with Efficient Scheduling and Optimizer Step Overlapping
标题：GreedySnake：通过高效的调度和优化步骤重叠加速SSD卸载LLM训练
链接：https://arxiv.org/abs/2512.17570
摘要：SSD卸载培训提供了一个实用和有前途的方法，使LLM培训具有成本效益。基于微批处理的梯度累积，本文介绍了GreedySnake，这是一种采用垂直调度的新SSD卸载训练系统，它在进入下一层之前执行一层的所有微批处理。与使用水平调度的现有系统（即，顺序执行微批处理），GreedySnake以较小的批处理大小实现了更高的训练吞吐量，使系统更接近roofline模型预测的理想场景。为了进一步缓解I/O瓶颈，GreedySnake将部分优化步骤与下一次迭代的前向传递重叠。在A100 GPU上的实验结果表明，GreedySnake实现了比ZeRO-Infinity更高的饱和训练吞吐量：对于GPT-65 B，在1个GPU上为1.96倍，在4个GPU上为1.93倍，对于GPT-175 B，在1个GPU上为2.53倍。该代码在https://github.com/npz7yyk/GreedySnake上开源
摘要：SSD-offloaded training offers a practical and promising approach to making LLM training cost-effective. Building on gradient accumulation with micro-batches, this paper introduces GreedySnake, a new SSD-offloaded training system that employs vertical scheduling, which executes all microbatches of a layer before proceeding to the next. Compared to existing systems that use horizontal scheduling (i.e., executing micro-batches sequentially), GreedySnake achieves higher training throughput with smaller batch sizes, bringing the system much closer to the ideal scenario predicted by the roofline model. To further mitigate the I/O bottleneck, GreedySnake overlaps part of the optimization step with the forward pass of the next iteration. Experimental results on A100 GPUs show that GreedySnake achieves saturated training throughput improvements over ZeRO-Infinity: 1.96x on 1 GPU and 1.93x on 4 GPUs for GPT-65B, and 2.53x on 1 GPU for GPT-175B. The code is open-sourced at https://github.com/npz7yyk/GreedySnake
【3】AdvJudge-Zero: Binary Decision Flips in LLM-as-a-Judge via Adversarial Control Tokens
标题：AdvJudge-Zero：通过对抗控制令牌在LLM担任法官中翻转二元决策
链接：https://arxiv.org/abs/2512.17375
摘要：奖励模型和LLM-as-a-Judge系统是RLHF、DPO和RLAIF等现代后训练管道的核心，它们提供标量反馈和二进制决策，指导模型选择和基于RL的微调。我们发现，这些判断系统表现出一个反复出现的漏洞：短序列的低困惑控制令牌可以翻转许多二进制评估从正确的“否”的判断不正确的“是”的判断，通过转向最后一层logit差距。这些控制令牌是策略模型在后期训练期间可能生成的模式，因此代表了现实的奖励黑客风险，而不是最坏情况下的对抗字符串。我们的方法，AdvJudge-Zero，使用模型的下一个令牌分布和波束搜索探索从头开始发现不同的控制令牌序列，我们的分析表明，诱导的隐藏状态扰动集中在一个低秩的“软模式”，与法官的拒绝方向相反。从经验上讲，当大型开放权重和专业判断模型在数学和推理基准上获得错误答案时，这些令牌会导致非常高的误报率。最后，我们证明了在少量控制令牌增强示例上进行基于LoRA的对抗性训练可以显着减少这些误报，同时保持评估质量。
摘要：Reward models and LLM-as-a-Judge systems are central to modern post-training pipelines such as RLHF, DPO, and RLAIF, where they provide scalar feedback and binary decisions that guide model selection and RL-based fine-tuning. We show that these judge systems exhibit a recurring vulnerability: short sequences of low-perplexity control tokens can flip many binary evaluations from correct ``No'' judgments to incorrect ``Yes'' judgments by steering the last-layer logit gap. These control tokens are patterns that a policy model could plausibly generate during post-training, and thus represent realistic reward-hacking risks rather than worst-case adversarial strings. Our method, AdvJudge-Zero, uses the model's next-token distribution and beam-search exploration to discover diverse control-token sequences from scratch, and our analysis shows that the induced hidden-state perturbations concentrate in a low-rank ``soft mode'' that is anti-aligned with the judge's refusal direction. Empirically, these tokens cause very high false positive rates when large open-weight and specialized judge models score incorrect answers on math and reasoning benchmarks. Finally, we show that LoRA-based adversarial training on small sets of control-token-augmented examples can markedly reduce these false positives while preserving evaluation quality.
【4】Verifiability-First Agents: Provable Observability and Lightweight Audit Agents for Controlling Autonomous LLM Systems
标题：可验证性优先代理：用于控制自治LLM系统的可证明可观察性和轻量级审计代理
链接：https://arxiv.org/abs/2512.17259
摘要：随着基于LLM的代理变得更加自主和多模式，确保它们保持可控，可审计和忠实于部署者意图变得至关重要。先前的基准测试测量了错位行为的倾向，并表明代理人的个性和工具访问显着影响错位。基于这些见解，我们提出了一个可验证性第一架构，（1）集成运行时认证的代理操作使用密码和符号的方法，（2）嵌入轻量级的审计代理，不断验证意图与行为使用约束推理，（3）执行挑战响应认证协议的高风险操作。我们介绍OPERA（可观察性，可证明执行，红队，证明），一个基准套件和评估协议，旨在衡量（i）未对准的可检测性，（ii）隐形策略下的检测时间，以及（iii）对抗性提示和角色注入的可验证性机制的弹性。我们的方法将评估重点从可能的不对准转移到如何快速可靠地检测和修复不对准。
摘要：As LLM-based agents grow more autonomous and multi-modal, ensuring they remain controllable, auditable, and faithful to deployer intent becomes critical. Prior benchmarks measured the propensity for misaligned behavior and showed that agent personalities and tool access significantly influence misalignment. Building on these insights, we propose a Verifiability-First architecture that (1) integrates run-time attestations of agent actions using cryptographic and symbolic methods, (2) embeds lightweight Audit Agents that continuously verify intent versus behavior using constrained reasoning, and (3) enforces challenge-response attestation protocols for high-risk operations. We introduce OPERA (Observability, Provable Execution, Red-team, Attestation), a benchmark suite and evaluation protocol designed to measure (i) detectability of misalignment, (ii) time to detection under stealthy strategies, and (iii) resilience of verifiability mechanisms to adversarial prompt and persona injection. Our approach shifts the evaluation focus from how likely misalignment is to how quickly and reliably misalignment can be detected and remediated.
【5】AlignDP: Hybrid Differential Privacy with Rarity-Aware Protection for LLMs
标题：AlignDP：针对LLM的混合差异隐私和稀有性感知保护
链接：https://arxiv.org/abs/2512.17251
备注：39th Conference on Neural Information Processing Systems (NeurIPS 2025) Workshop: LOCK-LLM Work-shop, NeurIPS 2025
摘要：大型语言模型面临着提取、提炼和未经授权的微调的风险。现有的防御措施使用水印或监控，但这些措施在泄漏后才起作用。我们设计了AlignDP，这是一种混合隐私锁，可以阻止数据接口处的知识转移。关键思想是将稀有字段和非稀有字段分开。罕见的领域是由PAC的不可逆性屏蔽，提供有效的零干扰本地DP。非稀有油气田通过RAPPOR私有化，根据当地DP提供无偏频率估计。全球聚合器强制执行组成和预算。这种双层设计隐藏了罕见事件，并为频繁事件添加了受控噪声。我们证明了PAC扩展到全局聚合的限制，给出了RAPPOR估计的界，并分析了效用权衡。一个玩具模拟证实了可行性：罕见的类别保持隐藏，频繁的类别恢复小的错误。
摘要：Large language models are exposed to risks of extraction, distillation, and unauthorized fine-tuning. Existing defenses use watermarking or monitoring, but these act after leakage. We design AlignDP, a hybrid privacy lock that blocks knowledge transfer at the data interface. The key idea is to separate rare and non-rare fields. Rare fields are shielded by PAC indistinguishability, giving effective zero-epsilon local DP. Non-rare fields are privatized with RAPPOR, giving unbiased frequency estimates under local DP. A global aggregator enforces composition and budget. This two-tier design hides rare events and adds controlled noise to frequent events. We prove limits of PAC extension to global aggregation, give bounds for RAPPOR estimates, and analyze utility trade-off. A toy simulation confirms feasibility: rare categories remain hidden, frequent categories are recovered with small error.
【6】Smoothing DiLoCo with Primal Averaging for Faster Training of LLMs
标题：利用Primal平均平滑DiLoCo，以更快地训练LLM
链接：https://arxiv.org/abs/2512.17131
摘要：我们提出了广义原始平均（GPA），Nesterov的方法在其原始平均公式的扩展，解决了最近的平均为基础的优化，如单工人DiLoCo和无调度（SF）在非分布式设置的关键限制。这两种最近的算法方法通过不同的平均策略提高了基本优化器（如AdamW）的性能。Schedule-Free显式地维护过去权重的统一平均值，而单工作者DiLoCo通过定期聚合轨迹（称为伪梯度）来执行隐式平均，以更新模型参数。然而，单工作者DiLoCo的周期平均引入了一个双循环结构，增加了其内存需求和超参数的数量。GPA通过解耦Nesterov原始平均公式中的插值常数克服了这些限制。这种解耦使GPA能够在每一步平滑地平均迭代，从而推广和改进单工作者DiLoCo。从经验来看，GPA的性能始终优于单工作线程DiLoCo，同时消除了双循环结构，简化了超参数调整，并将其内存开销减少到单个额外缓冲区。在Llama-160 M模型上，GPA在达到基线（AdamW）验证损失的步骤方面提供了24.22%的加速。同样，GPA在小批量和大批量设置上分别实现了12%和27%的加速，以达到AdamW在ImageNet ViT工作负载上的验证精度。此外，我们证明了，对于任何基础优化与遗憾界为$O（\sqrt{T}）$，其中$T$是迭代次数，GPA可以匹配或超过原来的优化的收敛保证，这取决于插值常数的选择。
摘要：We propose Generalized Primal Averaging (GPA), an extension of Nesterov's method in its primal averaging formulation that addresses key limitations of recent averaging-based optimizers such as single-worker DiLoCo and Schedule-Free (SF) in the non-distributed setting. These two recent algorithmic approaches improve the performance of base optimizers, such as AdamW, through different iterate averaging strategies. Schedule-Free explicitly maintains a uniform average of past weights, while single-worker DiLoCo performs implicit averaging by periodically aggregating trajectories, called pseudo-gradients, to update the model parameters. However, single-worker DiLoCo's periodic averaging introduces a two-loop structure, increasing its memory requirements and number of hyperparameters. GPA overcomes these limitations by decoupling the interpolation constant in the primal averaging formulation of Nesterov. This decoupling enables GPA to smoothly average iterates at every step, generalizing and improving upon single-worker DiLoCo. Empirically, GPA consistently outperforms single-worker DiLoCo while removing the two-loop structure, simplifying hyperparameter tuning, and reducing its memory overhead to a single additional buffer. On the Llama-160M model, GPA provides a 24.22% speedup in terms of steps to reach the baseline (AdamW's) validation loss. Likewise, GPA achieves speedups of 12% and 27% on small and large batch setups, respectively, to attain AdamW's validation accuracy on the ImageNet ViT workload. Furthermore, we prove that for any base optimizer with regret bounded by $O(\sqrt{T})$, where $T$ is the number of iterations, GPA can match or exceed the convergence guarantee of the original optimizer, depending on the choice of interpolation constants.
【7】A Women's Health Benchmark for Large Language Models
标题：大型语言模型的女性健康基准
链接：https://arxiv.org/abs/2512.17028
备注：15 pages, 6 Figures, 2 Tables
摘要：随着大型语言模型（LLM）成为数百万人健康信息的主要来源，它们在女性健康方面的准确性仍然没有得到严格的审查。我们介绍了妇女健康基准（WHB），第一个基准评估LLM性能专门在妇女的健康。我们的基准包括96个经过严格验证的模型树桩，涵盖五个医学专业（妇产科、急诊医学、初级保健、肿瘤学和神经学），三种查询类型（患者查询、临床医生查询和证据/政策查询）和八种错误类型（剂量/用药错误、缺失关键信息、过时的指南/治疗建议、不正确的治疗建议、不正确的事实信息，漏诊/不正确的鉴别诊断、漏诊紧急情况和不适当的建议）。我们评估了13个最先进的LLM，并揭示了惊人的差距：目前的模型显示，女性健康基准的失败率约为60%，不同专业和错误类型的表现差异很大。值得注意的是，模型普遍存在“错过紧急情况”指标的问题，而GPT-5等较新的模型在避免不适当的建议方面显示出显着的改进。我们的研究结果强调，人工智能聊天机器人还不能完全为女性健康提供可靠的建议。
摘要 ：As large language models (LLMs) become primary sources of health information for millions, their accuracy in women's health remains critically unexamined. We introduce the Women's Health Benchmark (WHB), the first benchmark evaluating LLM performance specifically in women's health. Our benchmark comprises 96 rigorously validated model stumps covering five medical specialties (obstetrics and gynecology, emergency medicine, primary care, oncology, and neurology), three query types (patient query, clinician query, and evidence/policy query), and eight error types (dosage/medication errors, missing critical information, outdated guidelines/treatment recommendations, incorrect treatment advice, incorrect factual information, missing/incorrect differential diagnosis, missed urgency, and inappropriate recommendations). We evaluated 13 state-of-the-art LLMs and revealed alarming gaps: current models show approximately 60\% failure rates on the women's health benchmark, with performance varying dramatically across specialties and error types. Notably, models universally struggle with "missed urgency" indicators, while newer models like GPT-5 show significant improvements in avoiding inappropriate recommendations. Our findings underscore that AI chatbots are not yet fully able of providing reliable advice in women's health.
【8】Turn-PPO: Turn-Level Advantage Estimation with PPO for Improved Multi-Turn RL in Agentic LLMs
标题：Turn-PPO：使用PPO进行回合级优势估计，以改进大型LLM中的多回合RL
链接：https://arxiv.org/abs/2512.17008
摘要：强化学习（RL）已经重新成为在现实世界环境中训练交互式LLM代理的自然方法。然而，直接将广泛使用的组相对策略优化（GRPO）算法应用于多回合任务暴露出显着的局限性，特别是在需要长时间推理的情况下。为了解决这些挑战，我们研究更稳定和有效的优势估计策略，特别是对于多回合设置。我们首先探索了邻近策略优化（PPO）作为一种替代方案，并发现它比GRPO更强大。为了进一步增强多回合场景中的PPO，我们引入了回合PPO，这是一种在回合级MDP公式上操作的变体，而不是常用的令牌级MDP。我们在WebShop和Sokoban数据集上的结果证明了turn-PPO的有效性，无论是否有长推理组件。
摘要：Reinforcement learning (RL) has re-emerged as a natural approach for training interactive LLM agents in real-world environments. However, directly applying the widely used Group Relative Policy Optimization (GRPO) algorithm to multi-turn tasks exposes notable limitations, particularly in scenarios requiring long-horizon reasoning. To address these challenges, we investigate more stable and effective advantage estimation strategies, especially for multi-turn settings. We first explore Proximal Policy Optimization (PPO) as an alternative and find it to be more robust than GRPO. To further enhance PPO in multi-turn scenarios, we introduce turn-PPO, a variant that operates on a turn-level MDP formulation, as opposed to the commonly used token-level MDP. Our results on the WebShop and Sokoban datasets demonstrate the effectiveness of turn-PPO, both with and without long reasoning components.
【9】Probing Scientific General Intelligence of LLMs with Scientist-Aligned Workflows
标题：通过科学一致的工作流程探索法学硕士的科学通用智能
链接：https://arxiv.org/abs/2512.16969
摘要：尽管科学人工智能取得了进展，但科学通用智能（SGI）的连贯框架-跨科学领域自主构思，调查和推理的能力-仍然缺乏。我们提出了一个基于实用探究模型（PIM：审议，概念，行动，感知）的可操作SGI定义，并通过四个科学家对齐的任务将其操作化：深入研究，想法生成，干/湿实验和实验推理。SGI-Bench由1，000多个专家策划的跨学科样本组成，灵感来自科学的125个大问题，能够对最先进的LLM进行系统评估。结果揭示了差距：尽管步骤级对齐，但在深度研究中精确匹配率低（10- 20%）;想法缺乏可行性和细节;在干实验中代码可执行性高，但执行结果准确性低;在湿协议中序列保真度低;以及持续的多模态比较推理挑战。我们进一步引入测试时强化学习（TTRL），它优化推理时检索增强的新颖性奖励，在没有参考答案的情况下增强假设新颖性。总之，我们基于PIM的定义，以工作流为中心的基准和经验见解为真正参与科学发现的AI系统奠定了基础。
摘要：Despite advances in scientific AI, a coherent framework for Scientific General Intelligence (SGI)-the ability to autonomously conceive, investigate, and reason across scientific domains-remains lacking. We present an operational SGI definition grounded in the Practical Inquiry Model (PIM: Deliberation, Conception, Action, Perception) and operationalize it via four scientist-aligned tasks: deep research, idea generation, dry/wet experiments, and experimental reasoning. SGI-Bench comprises over 1,000 expert-curated, cross-disciplinary samples inspired by Science's 125 Big Questions, enabling systematic evaluation of state-of-the-art LLMs. Results reveal gaps: low exact match (10--20%) in deep research despite step-level alignment; ideas lacking feasibility and detail; high code executability but low execution result accuracy in dry experiments; low sequence fidelity in wet protocols; and persistent multimodal comparative-reasoning challenges. We further introduce Test-Time Reinforcement Learning (TTRL), which optimizes retrieval-augmented novelty rewards at inference, enhancing hypothesis novelty without reference answer. Together, our PIM-grounded definition, workflow-centric benchmark, and empirical insights establish a foundation for AI systems that genuinely participate in scientific discovery.
【10】Compression is Routing: Reconstruction Error as an Intrinsic Signal for Modular Language Models
标题：压缩即路由：重建错误作为模块化语言模型的内在信号
链接：https://arxiv.org/abs/2512.16963
摘要：当前的大型语言模型（LLM）面临三个主要挑战：上下文长度限制，高推理成本和持续学习过程中的灾难性遗忘。虽然混合专家（MoE）架构减轻了这些冲突中的一些，但它们的路由机制通常依赖于显式训练的辅助分类器。这不仅增加了系统的复杂性，而且在处理混合域输入时往往缺乏可解释性。 基于“压缩即智能”的前提，本文提出了一种新的体系结构哲学：\textbf{``压缩即路由。''}我们训练了一个87 M参数的端到端Transformer自动编码器，实现了\textbf{64 x序列长度压缩}（将512个令牌压缩为8个潜在向量）。实验结果表明，该压缩器具有极强的域鉴别能力：在域内（代码）验证集上的重构精度达到99.47%，在半分布域上的重构精度急剧下降到47.76%（维基文本）;并进一步骤降到完全分布域（随机序列）上的仅\textbf{0.57\%}。 这种极端和系统的性能差异确立了重建误差作为\textbf{内在分布指纹}的有效性。在此基础上，我们提出，专家模块可以自动调度直接使用重建残差，而不需要显式的门控网络。这种机制提供了出色的可扩展性。此外，这种架构提供了一个新的视角“VRAM压缩”处理超长上下文。本报告旨在验证这种基础架构的物理有效性，为下一代可扩展模块化神经网络提供新的研究视角。
摘要 ：Current Large Language Models (LLMs) face three major challenges: context length limitations, high inference costs, and catastrophic forgetting during continual learning. While Mixture-of-Experts (MoE) architectures mitigate some of these conflicts, their routing mechanisms typically rely on explicitly trained auxiliary classifiers. This not only increases system complexity but also often lacks interpretability when handling mixed-domain inputs. Building upon the premise that ``Compression is Intelligence,'' this paper proposes a novel architectural philosophy: \textbf{``Compression is Routing.''} We trained an 87M-parameter end-to-end Transformer Autoencoder, achieving a \textbf{64x sequence length compression} (compressing 512 tokens into 8 latent vectors). Experimental results demonstrate that this compressor possesses extreme domain discriminative capability: it achieves a reconstruction accuracy of \textbf{99.47\%} on the in-domain (code) validation set; accuracy drops sharply to \textbf{47.76\%} on a semi-out-of-distribution domain (Wiki text); and further plummets to just \textbf{0.57\%} on a fully out-of-distribution domain (random sequences). This extreme and systematic performance discrepancy establishes the validity of reconstruction error as an \textbf{Intrinsic Distribution Fingerprint}. Based on this, we propose that expert modules can be automatically scheduled using reconstruction residuals directly, without the need for explicit gating networks. This mechanism offers excellent scalability. Furthermore, this architecture provides a new perspective on ``VRAM compression'' for handling ultra-long contexts. This report aims to verify the physical validity of this foundational architecture, offering a new research perspective for the next generation of scalable modular neural networks.
【11】MemoryGraft: Persistent Compromise of LLM Agents via Poisoned Experience Retrieval
标题：MemoryGraft：通过中毒经验检索持续损害LLM药物
链接：https://arxiv.org/abs/2512.16962
备注：14 pages, 1 figure, includes appendix
摘要：大型语言模型（LLM）代理越来越依赖于长期记忆和检索增强生成（RAG）来保持经验并改善未来的性能。虽然这种经验学习能力增强了代理人的自主性，但它引入了一个关键的、未经探索的攻击面，即，智能体的推理核心和它自己的过去之间的信任边界。在本文中，我们介绍MemoryGraft。这是一种新型的间接注入攻击，它不是通过立即越狱来损害代理的行为，而是通过将恶意的成功经验植入代理的长期记忆中。与传统的即时注入是短暂的，或标准的RAG中毒的目标事实知识，MemoryGraft利用代理的语义模仿启发式，这是复制模式的趋势检索成功的任务。我们证明，攻击者谁可以提供良性摄取水平的工件，代理在执行过程中读取可以诱导它构建一个中毒的RAG存储，其中一小部分恶意程序模板一起持久化良性的经验。当代理后来遇到语义相似的任务，联合检索词汇和嵌入相似性可靠地表面这些嫁接的记忆，和代理采用嵌入的不安全模式，导致持续的行为漂移跨会话。我们使用GPT-4 o在MetaGPT的DataInterpreter代理上验证MemoryGraft，发现少量中毒记录可以占良性工作负载上检索到的大部分经验，将基于经验的自我改进转化为隐形和持久妥协的载体。为了促进可重复性和未来的研究，我们的代码和评估数据可在https://github.com/Jacobhhy/Agent-Memory-Poisoning上获得。
摘要：Large Language Model (LLM) agents increasingly rely on long-term memory and Retrieval-Augmented Generation (RAG) to persist experiences and refine future performance. While this experience learning capability enhances agentic autonomy, it introduces a critical, unexplored attack surface, i.e., the trust boundary between an agent's reasoning core and its own past. In this paper, we introduce MemoryGraft. It is a novel indirect injection attack that compromises agent behavior not through immediate jailbreaks, but by implanting malicious successful experiences into the agent's long-term memory. Unlike traditional prompt injections that are transient, or standard RAG poisoning that targets factual knowledge, MemoryGraft exploits the agent's semantic imitation heuristic which is the tendency to replicate patterns from retrieved successful tasks. We demonstrate that an attacker who can supply benign ingestion-level artifacts that the agent reads during execution can induce it to construct a poisoned RAG store where a small set of malicious procedure templates is persisted alongside benign experiences. When the agent later encounters semantically similar tasks, union retrieval over lexical and embedding similarity reliably surfaces these grafted memories, and the agent adopts the embedded unsafe patterns, leading to persistent behavioral drift across sessions. We validate MemoryGraft on MetaGPT's DataInterpreter agent with GPT-4o and find that a small number of poisoned records can account for a large fraction of retrieved experiences on benign workloads, turning experience-based self-improvement into a vector for stealthy and durable compromise. To facilitate reproducibility and future research, our code and evaluation data are available at https://github.com/Jacobhhy/Agent-Memory-Poisoning.
Graph相关(图学习|图神经网络|图优化等)(7篇)
【1】Can You Hear Me Now? A Benchmark for Long-Range Graph Propagation
标题：你现在能听到我吗？长期图传播的基准
链接：https://arxiv.org/abs/2512.17762
摘要：有效地捕捉远程交互仍然是图神经网络（GNN）研究中一个基本但尚未解决的挑战，对于不同科学领域的应用至关重要。为了系统地解决这个问题，我们引入了ECHO（Evaluating Communication over long HOps），这是一种专门设计用于严格评估GNN处理长距离图传播能力的新基准。ECHO包括三个合成图任务，即单源最短路径，节点偏心率和图直径，每个任务都是在各种各样的结构上构建的，这些结构具有挑战性的拓扑结构旨在引入重要的信息瓶颈。ECHO还包括两个真实世界的数据集，ECHO-Charge和ECHO-Energy，它们分别定义了用于预测原子部分电荷和分子总能量的化学基准，并在密度泛函理论（DFT）水平上获得了参考计算。这两项任务本质上都依赖于捕获复杂的长程分子相互作用。我们对流行的GNN架构进行了广泛的基准测试，揭示了明显的性能差距，强调了真正的远程传播的难度，并强调了能够克服固有限制的设计选择。因此，ECHO为评估远程信息传播设定了一个新的标准，也为人工智能的科学需求提供了一个令人信服的例子。
摘要：Effectively capturing long-range interactions remains a fundamental yet unresolved challenge in graph neural network (GNN) research, critical for applications across diverse fields of science. To systematically address this, we introduce ECHO (Evaluating Communication over long HOps), a novel benchmark specifically designed to rigorously assess the capabilities of GNNs in handling very long-range graph propagation. ECHO includes three synthetic graph tasks, namely single-source shortest paths, node eccentricity, and graph diameter, each constructed over diverse and structurally challenging topologies intentionally designed to introduce significant information bottlenecks. ECHO also includes two real-world datasets, ECHO-Charge and ECHO-Energy, which define chemically grounded benchmarks for predicting atomic partial charges and molecular total energies, respectively, with reference computations obtained at the density functional theory (DFT) level. Both tasks inherently depend on capturing complex long-range molecular interactions. Our extensive benchmarking of popular GNN architectures reveals clear performance gaps, emphasizing the difficulty of true long-range propagation and highlighting design choices capable of overcoming inherent limitations. ECHO thereby sets a new standard for evaluating long-range information propagation, also providing a compelling example for its need in AI for science.
【2】A lightweight Spatial-Temporal Graph Neural Network for Long-term Time Series Forecasting
标题：用于长期时间序列预测的轻量级时空图神经网络
链接：https://arxiv.org/abs/2512.17453
备注：9 pages, 5 figures, 2 tables. Accepted for presentation at the 18th International Conference on Agents and Artificial Intelligence (ICAART 2026), Marbella, Spain
摘要：我们提出了Lite-STGNN，这是一种用于长期多变量预测的轻量级时空图神经网络，它将基于分解的时间建模与可学习的稀疏图结构相结合。时间模块应用趋势-季节分解，而空间模块使用低秩Top-$K$邻接学习和保守水平选通执行消息传递，从而实现增强强线性基线的空间校正。Lite-STGNN在四个基准数据集上达到了最先进的精度，范围高达720步，同时具有参数效率，并且比基于变换的方法更快地进行训练。消融的研究表明，空间模块产生4.6%的时间基线的改善，顶部-$K$增强局部性3.3%，学习邻接矩阵揭示特定领域的相互作用动态。因此，Lite-STGNN为长期多变量时间序列预测提供了一个紧凑，可解释和有效的框架。
摘要 ：We propose Lite-STGNN, a lightweight spatial-temporal graph neural network for long-term multivariate forecasting that integrates decomposition-based temporal modeling with learnable sparse graph structure. The temporal module applies trend-seasonal decomposition, while the spatial module performs message passing with low-rank Top-$K$ adjacency learning and conservative horizon-wise gating, enabling spatial corrections that enhance a strong linear baseline. Lite-STGNN achieves state-of-the-art accuracy on four benchmark datasets for horizons up to 720 steps, while being parameter-efficient and substantially faster to train than transformer-based methods. Ablation studies show that the spatial module yields 4.6% improvement over the temporal baseline, Top-$K$ enhances locality by 3.3%, and learned adjacency matrices reveal domain-specific interaction dynamics. Lite-STGNN thus offers a compact, interpretable, and efficient framework for long-term multivariate time series forecasting.
【3】Adaptive Graph Pruning with Sudden-Events Evaluation for Traffic Prediction using Online Semi-Decentralized ST-GNNs
标题：使用在线半分散ST-GNN进行流量预测的具有突发事件评估的自适应图修剪
链接：https://arxiv.org/abs/2512.17352
备注：19 pages, 6 figures, 5 tables, journal
摘要：时空图神经网络（ST-GNN）非常适合处理智能移动系统中地理分布传感器的高频数据流。然而，它们在分布式计算节点（cloudlets）的边缘部署由于相邻cloudlets之间重叠节点特征的重复传输而带来了大量的通信开销。为了解决这个问题，我们提出了一个自适应修剪算法，动态过滤冗余的邻居功能，同时保留最丰富的空间上下文预测。该算法根据最近的模型性能调整修剪率，允许每个小云专注于经历流量变化的区域，而不会影响准确性。此外，我们还介绍了突发事件预测准确性（SEPA），这是一种新的以事件为中心的指标，旨在衡量对流量减慢和恢复的响应，而标准错误指标往往会忽略这些指标。我们在两个大规模流量数据集PeMS-BAY和PeMSD 7-M上，在短期，中期和长期预测范围内使用传统FL，无服务器FL和Gossip Learning在线半分散设置评估我们的方法。实验表明，与标准度量相比，SEPA揭示了空间连通性在预测动态和不规则流量方面的真正价值。我们的自适应修剪算法保持了预测准确性，同时显著降低了所有在线半分散设置中的通信成本，表明可以在不影响对关键交通事件的响应的情况下减少通信。
摘要：Spatio-Temporal Graph Neural Networks (ST-GNNs) are well-suited for processing high-frequency data streams from geographically distributed sensors in smart mobility systems. However, their deployment at the edge across distributed compute nodes (cloudlets) createssubstantial communication overhead due to repeated transmission of overlapping node features between neighbouring cloudlets. To address this, we propose an adaptive pruning algorithm that dynamically filters redundant neighbour features while preserving the most informative spatial context for prediction. The algorithm adjusts pruning rates based on recent model performance, allowing each cloudlet to focus on regions experiencing traffic changes without compromising accuracy. Additionally, we introduce the Sudden Event Prediction Accuracy (SEPA), a novel event-focused metric designed to measure responsiveness to traffic slowdowns and recoveries, which are often missed by standard error metrics. We evaluate our approach in an online semi-decentralized setting with traditional FL, server-free FL, and Gossip Learning on two large-scale traffic datasets, PeMS-BAY and PeMSD7-M, across short-, mid-, and long-term prediction horizons. Experiments show that, in contrast to standard metrics, SEPA exposes the true value of spatial connectivity in predicting dynamic and irregular traffic. Our adaptive pruning algorithm maintains prediction accuracy while significantly lowering communication cost in all online semi-decentralized settings, demonstrating that communication can be reduced without compromising responsiveness to critical traffic events.
【4】CheXPO-v2: Preference Optimization for Chest X-ray VLMs with Knowledge Graph Consistency
标题：CheXPO-v2：具有知识图一致性的胸部X射线VLM偏好优化
链接：https://arxiv.org/abs/2512.17213
摘要：医学视觉语言模型（VLM）容易产生幻觉，影响临床可靠性。虽然像组相对策略优化（GRPO）这样的强化学习方法提供了一种低成本的对齐解决方案，但它们对稀疏的、基于结果的奖励的依赖无意中鼓励了模型“过度思考”--生成冗长、复杂和无法验证的思想链推理来证明答案的合理性。这种对结果的关注掩盖了事实错误，并带来了重大的安全风险。为了解决这个问题，我们提出了CheXPO-v2，这是一个新的对齐框架，从结果转向过程监督。我们的核心创新是知识图一致性奖励机制，由知识关系匹配驱动。通过明确地将推理步骤解析为结构化的“疾病、关系、解剖”三元组，我们提供了细粒度的监督，在原子水平上惩罚不连贯的逻辑和幻觉。将其与硬示例挖掘策略相结合，我们的方法在MIMIC-CXR-VQA等基准测试中的表现明显优于GRPO和最先进的模型。至关重要的是，CheXPO-v2仅使用5 k个样本就实现了最先进的准确性，在产生临床合理和可验证的推理的同时，展示了卓越的数据效率。该项目的源代码可在https://github.com/ecoxial2007/CheX-Phi4MM上公开获得。
摘要：Medical Vision-Language Models (VLMs) are prone to hallucinations, compromising clinical reliability. While reinforcement learning methods like Group Relative Policy Optimization (GRPO) offer a low-cost alignment solution, their reliance on sparse, outcome-based rewards inadvertently encourages models to "overthink" -- generating verbose, convoluted, and unverifiable Chain-of-Thought reasoning to justify answers. This focus on outcomes obscures factual errors and poses significant safety risks. To address this, we propose CheXPO-v2, a novel alignment framework that shifts from outcome to process supervision. Our core innovation is a Knowledge Graph Consistency Reward mechanism driven by Entity-Relation Matching. By explicitly parsing reasoning steps into structured "Disease, Relation, Anatomy" triplets, we provide fine-grained supervision that penalizes incoherent logic and hallucinations at the atomic level. Integrating this with a hard-example mining strategy, our approach significantly outperforms GRPO and state-of-the-art models on benchmarks like MIMIC-CXR-VQA. Crucially, CheXPO-v2 achieves new state-of-the-art accuracy using only 5k samples, demonstrating exceptional data efficiency while producing clinically sound and verifiable reasoning. The project source code is publicly available at: https://github.com/ecoxial2007/CheX-Phi4MM.
【5】Distributed Learning in Markovian Restless Bandits over Interference Graphs for Stable Spectrum Sharing
标题：基于干扰图的马尔科夫不安Bandits分布式学习以实现稳定频谱共享
链接：https://arxiv.org/abs/2512.17161
备注：13 pages, 10 figures
摘要：我们研究分布式学习的频谱接入和共享之间的多个认知通信实体，如细胞，子网，或认知无线电用户（统称为细胞），在通信受限的无线网络干扰图建模。我们的目标是实现一个全球稳定和干扰感知的信道分配。通过广义Gale-Shapley多对一匹配来定义稳定性，这是无线资源分配中的一个完善的解决方案概念。我们考虑无线网络，其中L个小区共享S个正交信道，并且不能同时使用与它们的邻居相同的信道。每个信道作为一个未知的不稳定的马尔可夫过程与细胞相关的奖励，使这第一个工作，建立全球盖尔-Shapley稳定的信道分配在一个随机的，随时间变化的不稳定的环境。为了应对这一挑战，我们开发了SMILE（Stable Multi-matching with Interference-aware LEarning），这是一种通信效率高的分布式学习算法，将不安分的强盗学习与图约束协调相结合。SMILE使细胞能够分布式地平衡未知通道的探索与学习信息的利用。我们证明了SMILE收敛到最优的稳定分配，并实现对数遗憾相对于一个精灵与预期效用的充分知识。仿真验证了理论保证，并证明了SMILE的鲁棒性，可扩展性和效率在不同的频谱共享方案。
摘要 ：We study distributed learning for spectrum access and sharing among multiple cognitive communication entities, such as cells, subnetworks, or cognitive radio users (collectively referred to as cells), in communication-constrained wireless networks modeled by interference graphs. Our goal is to achieve a globally stable and interference-aware channel allocation. Stability is defined through a generalized Gale-Shapley multi-to-one matching, a well-established solution concept in wireless resource allocation. We consider wireless networks where L cells share S orthogonal channels and cannot simultaneously use the same channel as their neighbors. Each channel evolves as an unknown restless Markov process with cell-dependent rewards, making this the first work to establish global Gale-Shapley stability for channel allocation in a stochastic, temporally varying restless environment. To address this challenge, we develop SMILE (Stable Multi-matching with Interference-aware LEarning), a communication-efficient distributed learning algorithm that integrates restless bandit learning with graph-constrained coordination. SMILE enables cells to distributedly balance exploration of unknown channels with exploitation of learned information. We prove that SMILE converges to the optimal stable allocation and achieves logarithmic regret relative to a genie with full knowledge of expected utilities. Simulations validate the theoretical guarantees and demonstrate SMILE's robustness, scalability, and efficiency across diverse spectrum-sharing scenarios.
【6】Systemic Risk Radar: A Multi-Layer Graph Framework for Early Market Crash Warning
标题：系统性风险雷达：早期市场崩溃预警的多层图形框架
链接：https://arxiv.org/abs/2512.17185
备注：Preprint
摘要：当结构性脆弱性在各个部门、市场和投资者行为中累积时，金融危机就会出现。预测这些系统性转变具有挑战性，因为它们来自市场参与者之间不断变化的互动，而不仅仅是孤立的价格变动。我们提出了系统性风险雷达（SRR），一个框架，模型金融市场的多层图，以检测系统性脆弱性和崩溃政权过渡的早期迹象。 我们在三个主要危机中评估SRR：网络泡沫崩溃，全球金融危机和COVID-19冲击。我们的实验比较了快照GNN，简化的时间GNN原型和标准基线（逻辑回归和随机森林）。结果表明，结构网络信息提供了有用的预警信号相比，基于特征的模型。 这种基于相关性的SRR实例表明，图形衍生的功能捕捉有意义的变化，在压力事件期间的市场结构。这些发现激励人们通过额外的图形层（行业/因素暴露、情绪）和更具表达力的时态架构（LSTM/GRU或Transformer编码器）来扩展SRR，以更好地处理不同的危机类型。
摘要：Financial crises emerge when structural vulnerabilities accumulate across sectors, markets, and investor behavior. Predicting these systemic transitions is challenging because they arise from evolving interactions between market participants, not isolated price movements alone. We present Systemic Risk Radar (SRR), a framework that models financial markets as multi-layer graphs to detect early signs of systemic fragility and crash-regime transitions. We evaluate SRR across three major crises: the Dot-com crash, the Global Financial Crisis, and the COVID-19 shock. Our experiments compare snapshot GNNs, a simplified temporal GNN prototype, and standard baselines (logistic regression and Random Forest). Results show that structural network information provides useful early-warning signals compared to feature-based models alone. This correlation-based instantiation of SRR demonstrates that graph-derived features capture meaningful changes in market structure during stress events. The findings motivate extending SRR with additional graph layers (sector/factor exposure, sentiment) and more expressive temporal architectures (LSTM/GRU or Transformer encoders) to better handle diverse crisis types.
【7】Graph Attention Networks for Detecting Epilepsy from EEG Signals Using Accessible Hardware in Low-Resource Settings
标题：在低资源环境中使用可用硬件从脑电信号检测癫痫的图形注意力网络
链接：https://arxiv.org/abs/2507.15118
摘要：目标：由于神经科医生稀缺和诊断工具昂贵，癫痫在低收入国家仍然诊断不足。我们提出了一个基于图形的深度学习框架，通过低成本的脑电描记术（EEG）硬件检测癫痫，并对来自尼日利亚和几内亚比绍的录音进行了测试。我们的重点是公平，可访问的自动评估和可解释性，以阐明癫痫生物标志物。研究方法：我们将脑电信号建模为时空图，对其进行分类，并使用图注意力网络（GAT）识别通道间关系和时间动态。为了强调连接性生物标志物，我们采用固有的以节点为中心的GAT来分析边缘。我们还为低保真录音设计了信号预处理，并在Google Colab上训练了一个轻量级的GAT架构，并部署在RaspberryPi设备上。结果如下：该方法实现了有前途的分类性能，在多个会话的准确性和鲁棒性方面优于基于随机森林和图卷积网络的标准分类器，但也突出了额颞区的特定连接。结论：研究结果强调了GAT为服务不足地区的癫痫提供有见地和可扩展的诊断支持的潜力，为负担得起和可获得的神经诊断工具铺平了道路。
摘要：Goal: Epilepsy remains under-diagnosed in low-income countries due to scarce neurologists and costly diagnostic tools. We propose a graph-based deep learning framework to detect epilepsy from low-cost Electroencephalography (EEG) hardware, tested on recordings from Nigeria and Guinea-Bissau. Our focus is on fair, accessible automatic assessment and explainability to shed light on epilepsy biomarkers. Methods: We model EEG signals as spatio-temporal graphs, classify them, and identify interchannel relationships and temporal dynamics using graph attention networks (GAT). To emphasize connectivity biomarkers, we adapt the inherently node-focused GAT to analyze edges. We also designed signal preprocessing for low-fidelity recordings and a lightweight GAT architecture trained on Google Colab and deployed on RaspberryPi devices. Results: The approach achieves promising classification performance, outperforming a standard classifier based on random forest and graph convolutional networks in terms of accuracy and robustness over multiple sessions, but also highlighting specific connections in the fronto-temporal region. Conclusions: The results highlight the potential of GATs to provide insightful and scalable diagnostic support for epilepsy in underserved regions, paving the way for affordable and accessible neurodiagnostic tools.
Transformer(2篇)
【1】Spatially-informed transformers: Injecting geostatistical covariance biases into self-attention for spatio-temporal forecasting
标题：空间信息转换器：将地统计协方差偏差注入时空预测的自我注意力
链接：https://arxiv.org/abs/2512.17696
摘要：高维时空过程的建模在经典地质统计学的概率严格性和深度学习的灵活、高容量表示之间呈现出一种基本的二分法。虽然高斯过程提供了理论上的一致性和精确的不确定性量化，但其禁止的计算缩放使其对于大规模传感器网络不切实际。相反，现代Transformer架构擅长序列建模，但本质上缺乏几何归纳偏差，将空间传感器视为置换不变的令牌，而没有对距离的原生理解。在这项工作中，我们提出了一个空间信息的Transformer，一个混合架构，直接注入到自我注意力机制通过一个可学习的协方差内核的地质统计感应偏差。通过将注意力结构正式分解为一个固定的物理先验和一个非固定的数据驱动的残差，我们施加了一个软拓扑约束，有利于空间上接近的相互作用，同时保留模拟复杂动态的能力。我们展示了“深度变异”现象，其中网络通过反向传播成功地端到端恢复了底层过程的真实空间衰减参数。在合成高斯随机场和现实世界的流量基准上进行的大量实验证实，我们的方法优于最先进的图神经网络。此外，严格的统计验证证实，所提出的方法不仅提供了卓越的预测准确性，而且还提供了校准良好的概率预测，有效地弥合了物理感知建模和数据驱动学习之间的差距。
摘要：The modeling of high-dimensional spatio-temporal processes presents a fundamental dichotomy between the probabilistic rigor of classical geostatistics and the flexible, high-capacity representations of deep learning. While Gaussian processes offer theoretical consistency and exact uncertainty quantification, their prohibitive computational scaling renders them impractical for massive sensor networks. Conversely, modern transformer architectures excel at sequence modeling but inherently lack a geometric inductive bias, treating spatial sensors as permutation-invariant tokens without a native understanding of distance. In this work, we propose a spatially-informed transformer, a hybrid architecture that injects a geostatistical inductive bias directly into the self-attention mechanism via a learnable covariance kernel. By formally decomposing the attention structure into a stationary physical prior and a non-stationary data-driven residual, we impose a soft topological constraint that favors spatially proximal interactions while retaining the capacity to model complex dynamics. We demonstrate the phenomenon of ``Deep Variography'', where the network successfully recovers the true spatial decay parameters of the underlying process end-to-end via backpropagation. Extensive experiments on synthetic Gaussian random fields and real-world traffic benchmarks confirm that our method outperforms state-of-the-art graph neural networks. Furthermore, rigorous statistical validation confirms that the proposed method delivers not only superior predictive accuracy but also well-calibrated probabilistic forecasts, effectively bridging the gap between physics-aware modeling and data-driven learning.
【2】Colormap-Enhanced Vision Transformers for MRI-Based Multiclass (4-Class) Alzheimer's Disease Classification
标题：用于基于MRI的多类（4类）阿尔茨海默病分类的Colormap增强视觉变换器
链接：https://arxiv.org/abs/2512.16964
备注：12 pages, 4 figures
摘要：磁共振成像（MRI）在阿尔茨海默病（AD）的早期诊断和监测中起着关键作用。然而，大脑MRI扫描中的细微结构变化通常会对传统的深度学习模型提出挑战，以有效地提取区分特征。在这项工作中，我们提出了PseudoColorViT-Alz，一个色图增强的Vision Transformer框架，旨在利用伪彩色表示的MRI图像，以改善阿尔茨海默病的分类。通过将色图变换与Vision Transformers的全局特征学习功能相结合，我们的方法放大了在标准灰度MRI扫描中被抑制的解剖纹理和对比度线索。 我们使用四类分类设置（非痴呆，中度痴呆，轻度痴呆和非常轻度痴呆）在OASIS-1数据集上评估PseudoColorViT-Alz。我们的模型实现了99.79%的最新准确度，AUC为100%，超过了最近2024- 2025年方法的性能，包括基于CNN和Siamese网络的方法，其准确度范围为96.1%至99.68%。这些结果表明，伪彩色增强结合Vision Transformers可以显著增强基于MRI的阿尔茨海默病分类。PseudoColorViT-Alz提供了一个强大的和可解释的框架，优于目前的方法，提供了一个有前途的工具，以支持临床决策和早期检测阿尔茨海默病。
摘要：Magnetic Resonance Imaging (MRI) plays a pivotal role in the early diagnosis and monitoring of Alzheimer's disease (AD). However, the subtle structural variations in brain MRI scans often pose challenges for conventional deep learning models to extract discriminative features effectively. In this work, we propose PseudoColorViT-Alz, a colormap-enhanced Vision Transformer framework designed to leverage pseudo-color representations of MRI images for improved Alzheimer's disease classification. By combining colormap transformations with the global feature learning capabilities of Vision Transformers, our method amplifies anatomical texture and contrast cues that are otherwise subdued in standard grayscale MRI scans. We evaluate PseudoColorViT-Alz on the OASIS-1 dataset using a four-class classification setup (non-demented, moderate dementia, mild dementia, and very mild dementia). Our model achieves a state-of-the-art accuracy of 99.79% with an AUC of 100%, surpassing the performance of recent 2024--2025 methods, including CNN-based and Siamese-network approaches, which reported accuracies ranging from 96.1% to 99.68%. These results demonstrate that pseudo-color augmentation combined with Vision Transformers can significantly enhance MRI-based Alzheimer's disease classification. PseudoColorViT-Alz offers a robust and interpretable framework that outperforms current methods, providing a promising tool to support clinical decision-making and early detection of Alzheimer's disease.
GAN|对抗|攻击|生成相关(3篇)
【1】RadarGen: Automotive Radar Point Cloud Generation from Cameras
标题：RadarGen：从摄像机生成汽车雷达点云
链接：https://arxiv.org/abs/2512.17897
备注：Project page: https://radargen.github.io/
摘要：我们提出了RadarGen，一个扩散模型，用于从多视图相机图像合成逼真的汽车雷达点云。RadarGen通过以鸟瞰图形式表示雷达测量值，将空间结构与雷达截面（RCS）和多普勒属性一起编码，从而将有效的图像潜在扩散适应于雷达域。一个轻量级的恢复步骤从生成的地图重建点云。为了更好地将生成与视觉场景对齐，RadarGen结合了从预训练的基础模型中提取的BEV对齐的深度，语义和运动线索，这些线索将随机生成过程引导到物理上合理的雷达模式。对图像的调节使得该方法在原则上与现有的视觉数据集和仿真框架广泛兼容，为多模态生成仿真提供了可扩展的方向。对大规模驾驶数据的评估表明，RadarGen捕获了特征雷达测量分布，并缩小了与在真实数据上训练的感知模型的差距，标志着跨传感模式的统一生成仿真迈出了一步。
摘要：We present RadarGen, a diffusion model for synthesizing realistic automotive radar point clouds from multi-view camera imagery. RadarGen adapts efficient image-latent diffusion to the radar domain by representing radar measurements in bird's-eye-view form that encodes spatial structure together with radar cross section (RCS) and Doppler attributes. A lightweight recovery step reconstructs point clouds from the generated maps. To better align generation with the visual scene, RadarGen incorporates BEV-aligned depth, semantic, and motion cues extracted from pretrained foundation models, which guide the stochastic generation process toward physically plausible radar patterns. Conditioning on images makes the approach broadly compatible, in principle, with existing visual datasets and simulation frameworks, offering a scalable direction for multimodal generative simulation. Evaluations on large-scale driving data show that RadarGen captures characteristic radar measurement distributions and reduces the gap to perception models trained on real data, marking a step toward unified generative simulation across sensing modalities.
【2】SWE-Bench++: A Framework for the Scalable Generation of Software Engineering Benchmarks from Open-Source Repositories
标题：SWE-Bench++：从开源存储库可扩展生成软件工程基准的框架
链接：https://arxiv.org/abs/2512.17419
摘要：像SWE-bench这样的基准测试已经标准化了大型语言模型（LLM）在存储库级软件工程任务上的评估。然而，这些努力仍然受到手动策展、静态数据集以及专注于基于Python的错误修复的限制。我们介绍SWE-Bench++，这是一个自动化框架，可以从开源GitHub项目中生成存储库级别的编码任务。与合成方法不同，我们的管道收集实时拉取请求，以涵盖11种语言的错误修复和功能请求。SWE-Bench++通过四个阶段将GitHub拉取请求（PR）转换为可重复的、基于执行的任务：程序化采购、环境合成、测试oracle提取和质量保证。最后一个提示引导的轨迹合成步骤将强模型失败的实例转换为训练轨迹。我们的初始基准测试由来自11种语言的3，971个存储库的11，133个实例组成。在该基准测试的1，782个子集上，目前最强的模型表现如下：Claude-Sonnet-4.5达到36.20%pass@10，gpt-5-2025-08-07达到34.57%，gemini/gemini-2.5-pro达到24.92%，gpt-4 o达到16.89%。我们进一步证明了我们的数据集的实用性，通过对SWE-Bench++实例进行微调，可以在SWE-Bench多语言基准测试中获得可衡量的改进。SWE-Bench++提供了一个可扩展的多语言基准测试，用于评估和改进存储库级别的代码生成。
摘要：Benchmarks like SWE-bench have standardized the evaluation of Large Language Models (LLMs) on repository-level software engineering tasks. However, these efforts remain limited by manual curation, static datasets, and a focus on Python-based bug fixes. We introduce SWE-Bench++, an automated framework that generates repository-level coding tasks from open-source GitHub projects. Unlike synthetic approaches, our pipeline harvests live pull requests to cover both bug fixes and feature requests across 11 languages. SWE-Bench++ turns GitHub pull requests (PRs) into reproducible, execution-based tasks via four stages: programmatic sourcing, environment synthesis, test oracle extraction, and quality assurance. A final hint-guided trajectory synthesis step converts instances that strong models fail on into training trajectories. Our initial benchmark consists of 11,133 instances from 3,971 repositories across 11 languages. On a subset of 1,782 instances of this benchmark, today's strongest models perform as follows: claude-sonnet-4.5 achieves 36.20% pass@10, gpt-5-2025-08-07 34.57%, gemini/gemini-2.5-pro 24.92%, and gpt-4o 16.89%. We further demonstrate the utility of our dataset by showing that fine-tuning on SWE-Bench++ instances yields measurable improvements on the SWE-bench Multilingual benchmark. SWE-Bench++ provides a scalable, multilingual benchmark for evaluating and improving repository-level code generation.
【3】Biosecurity-Aware AI: Agentic Risk Auditing of Soft Prompt Attacks on ESM-Based Variant Predictors
标题：具有生物安全意识的人工智能：对基于ESM的变体预测器的软提示攻击的量化风险审计
链接：https://arxiv.org/abs/2512.17146
摘要：基因组基础模型（GFM），如进化尺度模型（ESM），在变异效应预测方面取得了显着的成功。然而，它们在对抗性操纵下的安全性和鲁棒性在很大程度上仍未得到探索。为了解决这一差距，我们引入了安全基因组评估器（SAGE），这是一个用于审计GFM对抗性漏洞的代理框架。SAGE通过可解释的自动化风险审计循环发挥作用。它注入软提示扰动，跨训练检查点监控模型行为，计算AUROC和AUPR等风险指标，并生成具有基于大型语言模型的叙述性解释的结构化报告。这种代理过程可以在不修改底层模型的情况下连续评估嵌入空间的鲁棒性。使用SAGE，我们发现，即使是最先进的GFM，如ESM 2是敏感的有针对性的软提示攻击，导致可测量的性能下降。这些发现揭示了基因组基础模型中先前隐藏的关键漏洞，表明代理风险审计在确保临床变异解读等生物医学应用方面的重要性。
摘要：Genomic Foundation Models (GFMs), such as Evolutionary Scale Modeling (ESM), have demonstrated remarkable success in variant effect prediction. However, their security and robustness under adversarial manipulation remain largely unexplored. To address this gap, we introduce the Secure Agentic Genomic Evaluator (SAGE), an agentic framework for auditing the adversarial vulnerabilities of GFMs. SAGE functions through an interpretable and automated risk auditing loop. It injects soft prompt perturbations, monitors model behavior across training checkpoints, computes risk metrics such as AUROC and AUPR, and generates structured reports with large language model-based narrative explanations. This agentic process enables continuous evaluation of embedding-space robustness without modifying the underlying model. Using SAGE, we find that even state-of-the-art GFMs like ESM2 are sensitive to targeted soft prompt attacks, resulting in measurable performance degradation. These findings reveal critical and previously hidden vulnerabilities in genomic foundation models, showing the importance of agentic risk auditing in securing biomedical applications such as clinical variant interpretation.
半/弱/无/有监督|不确定性|主动学习(3篇)
【1】Re-Depth Anything: Test-Time Depth Refinement via Self-Supervised Re-lighting
标题：重新深度任何内容：通过自我监督重新照明进行测试时深度细化
链接：https://arxiv.org/abs/2512.17908
摘要：单目深度估计仍然具有挑战性，因为最近的基础模型，如Depth Anything V2（DA-V2），与远离训练分布的真实世界图像斗争。我们引入了Re-Depth Anything，这是一个测试时自我监督框架，通过将DA-V2与大规模2D扩散模型的强大先验融合来弥合这一领域差距。我们的方法直接对输入图像进行无标签细化，通过重新照明预测的深度图和增强输入。这种重新合成方法通过在具有分数蒸馏采样（SDS）的新的生成上下文中利用阴影恢复形状（SfS）线索来取代经典的光度重建。为了防止优化崩溃，我们的框架采用了有针对性的优化策略：而不是直接优化深度或微调完整模型，我们冻结编码器，只更新中间嵌入，同时微调解码器。在不同的基准测试中，Re-Depth Anything在深度准确性和真实性方面比DA-V2有了很大的提高，通过增强几何推理展示了自我监督的新途径。
摘要：Monocular depth estimation remains challenging as recent foundation models, such as Depth Anything V2 (DA-V2), struggle with real-world images that are far from the training distribution. We introduce Re-Depth Anything, a test-time self-supervision framework that bridges this domain gap by fusing DA-V2 with the powerful priors of large-scale 2D diffusion models. Our method performs label-free refinement directly on the input image by re-lighting predicted depth maps and augmenting the input. This re-synthesis method replaces classical photometric reconstruction by leveraging shape from shading (SfS) cues in a new, generative context with Score Distillation Sampling (SDS). To prevent optimization collapse, our framework employs a targeted optimization strategy: rather than optimizing depth directly or fine-tuning the full model, we freeze the encoder and only update intermediate embeddings while also fine-tuning the decoder. Across diverse benchmarks, Re-Depth Anything yields substantial gains in depth accuracy and realism over the DA-V2, showcasing new avenues for self-supervision by augmenting geometric reasoning.
【2】MedNeXt-v2: Scaling 3D ConvNeXts for Large-Scale Supervised Representation Learning in Medical Image Segmentation
标题：MedNeXt-v2：缩放3D ConvNeXts，用于医学图像分割中的大规模监督表示学习
链接：https://arxiv.org/abs/2512.17774
摘要：大规模监督预训练正在快速重塑3D医学图像分割。然而，现有的努力主要集中在增加数据集的大小，忽略了骨干网络是否是一个有效的表示学习器的问题。在这项工作中，我们通过重新审视基于ConvNeXt的体积分割架构并引入MedNeXt-v2来解决这一差距，MedNeXt-v2是一种复合缩放的3D ConvNeXt，它利用改进的微架构和数据缩放来提供最先进的性能。首先，我们证明了在大规模预训练管道中常规使用的主干通常是次优的。随后，我们在扩展之前使用了全面的骨干基准测试，并证明了更强的从头开始的性能可靠地预测了预训练后更强的下游性能。在这些发现的指导下，我们结合了一个3D全局响应归一化模块，并使用深度，宽度和上下文缩放来改进我们的架构，以实现有效的表示学习。我们在18 k CT体积上预训练MedNeXt-v2，并在六个具有挑战性的CT和MR基准（144个结构）上进行微调时展示了最先进的性能，显示出超过七个公开发布的预训练模型的一致收益。除了改进之外，我们对这些模型的基准测试还表明，更强的主干在类似数据上会产生更好的结果，表示缩放对病理分割有不成比例的好处，并且一旦应用完全微调，特定于模态的预训练提供的好处可以忽略不计。总之，我们的研究结果将MedNeXt-v2确立为3D医学图像分割中大规模监督表示学习的强大支柱。我们的代码和预训练模型可在官方nnUNet存储库中获得：https://www.github.com/MIC-DKFZ/nnUNet
摘要：Large-scale supervised pretraining is rapidly reshaping 3D medical image segmentation. However, existing efforts focus primarily on increasing dataset size and overlook the question of whether the backbone network is an effective representation learner at scale. In this work, we address this gap by revisiting ConvNeXt-based architectures for volumetric segmentation and introducing MedNeXt-v2, a compound-scaled 3D ConvNeXt that leverages improved micro-architecture and data scaling to deliver state-of-the-art performance. First, we show that routinely used backbones in large-scale pretraining pipelines are often suboptimal. Subsequently, we use comprehensive backbone benchmarking prior to scaling and demonstrate that stronger from scratch performance reliably predicts stronger downstream performance after pretraining. Guided by these findings, we incorporate a 3D Global Response Normalization module and use depth, width, and context scaling to improve our architecture for effective representation learning. We pretrain MedNeXt-v2 on 18k CT volumes and demonstrate state-of-the-art performance when fine-tuning across six challenging CT and MR benchmarks (144 structures), showing consistent gains over seven publicly released pretrained models. Beyond improvements, our benchmarking of these models also reveals that stronger backbones yield better results on similar data, representation scaling disproportionately benefits pathological segmentation, and that modality-specific pretraining offers negligible benefit once full finetuning is applied. In conclusion, our results establish MedNeXt-v2 as a strong backbone for large-scale supervised representation learning in 3D Medical Image Segmentation. Our code and pretrained models are made available with the official nnUNet repository at: https://www.github.com/MIC-DKFZ/nnUNet
【3】Imputation Uncertainty in Interpretable Machine Learning Methods
标题：可解释机器学习方法中的归因不确定性
链接：https://arxiv.org/abs/2512.17689
备注：19 pages, 15 Figures, accepted at conference: IJCAI 2025 Workshop on Explainable Artificial Intelligence (Montreal, Canada)
摘要：在真实数据中，缺失值经常发生，这影响了可解释机器学习（IML）方法的解释。最近的工作考虑了偏差，并表明模型解释可能不同的插补方法，而忽略了额外的插补不确定性及其对方差和置信区间的影响。因此，我们比较了不同的填补方法对IML方法的置信区间覆盖概率的影响，排列特征重要性，部分依赖图和Shapley值。我们发现，单一填补导致低估的方差，在大多数情况下，只有多重填补接近名义覆盖。
摘要：In real data, missing values occur frequently, which affects the interpretation with interpretable machine learning (IML) methods. Recent work considers bias and shows that model explanations may differ between imputation methods, while ignoring additional imputation uncertainty and its influence on variance and confidence intervals. We therefore compare the effects of different imputation methods on the confidence interval coverage probabilities of the IML methods permutation feature importance, partial dependence plots and Shapley values. We show that single imputation leads to underestimation of variance and that, in most cases, only multiple imputation is close to nominal coverage.
迁移|Zero/Few/One-Shot|自适应(6篇)
【1】Easy Adaptation: An Efficient Task-Specific Knowledge Injection Method for Large Models in Resource-Constrained Environments
标题：轻松适应：资源受限环境中大型模型的高效任务特定知识注入方法
链接：https://arxiv.org/abs/2512.17771
摘要：虽然巨大的参数规模赋予大型模型（LM）无与伦比的性能，但它也限制了它们在特定任务中的适应性。参数有效的微调（PEFT）已成为一个关键的方法，有效地适应LM的下游任务的范围。然而，现有的PEFT方法面临两个主要挑战：（1）高资源成本。尽管PEFT方法与完全微调相比显著减少了资源需求，但它仍然需要大量的时间和内存，使得其在资源受限的环境中不切实际。(2)参数依赖性。PEFT方法严重依赖于更新与LM相关联的参数的子集，以结合特定于任务的知识。然而，由于LMs领域的竞争日益激烈，许多公司都对其领先的模型采用了闭源策略，仅通过应用程序编程接口（API）提供访问。然而，由于LM的微调过程极其缓慢，因此费用通常是成本高昂的并且难以维持。即使小模型的性能一般比LM差得多，它们也可以在特定分布上获得更好的结果，同时只需要最少的资源。基于这一认识，我们提出了Easy Adaptation（EA），它设计了特定的小模型（SSM）来补充LM的欠拟合数据分布。大量的实验表明，EA匹配PEFT的性能在不同的任务，而不访问LM参数，只需要最少的资源。
摘要：While the enormous parameter scale endows Large Models (LMs) with unparalleled performance, it also limits their adaptability across specific tasks. Parameter-Efficient Fine-Tuning (PEFT) has emerged as a critical approach for effectively adapting LMs to a diverse range of downstream tasks. However, existing PEFT methods face two primary challenges: (1) High resource cost. Although PEFT methods significantly reduce resource demands compared to full fine-tuning, it still requires substantial time and memory, making it impractical in resource-constrained environments. (2) Parameter dependency. PEFT methods heavily rely on updating a subset of parameters associated with LMs to incorporate task-specific knowledge. Yet, due to increasing competition in the LMs landscape, many companies have adopted closed-source policies for their leading models, offering access only via Application Programming Interface (APIs). Whereas, the expense is often cost-prohibitive and difficult to sustain, as the fine-tuning process of LMs is extremely slow. Even if small models perform far worse than LMs in general, they can achieve superior results on particular distributions while requiring only minimal resources. Motivated by this insight, we propose Easy Adaptation (EA), which designs Specific Small Models (SSMs) to complement the underfitted data distribution for LMs. Extensive experiments show that EA matches the performance of PEFT on diverse tasks without accessing LM parameters, and requires only minimal resources.
【2】Mitigating Forgetting in Low Rank Adaptation
标题：减少低级别适应中的遗忘
链接：https://arxiv.org/abs/2512.17720
摘要：参数高效的微调方法，如低秩自适应（LoRA），使大型预训练模型能够快速专业化到不同的下游应用。然而，这个过程往往会导致灾难性的遗忘模型的先验知识。我们使用LaLoRA解决这个问题，LaLoRA是一种权重空间正则化技术，将拉普拉斯近似应用于低秩自适应。我们的方法估计模型对每个参数的置信度，并限制高曲率方向的更新，保留先验知识，同时实现有效的目标域学习。通过仅将拉普拉斯近似应用于LoRA权重，该方法保持轻量。我们通过微调Llama模型进行数学推理来评估LaLoRA，并展示了改进的学习-遗忘权衡，这可以通过该方法的正则化强度直接控制。我们进一步探讨了不同的损失景观曲率近似估计参数的置信度，分析用于拉普拉斯近似的数据的效果，并研究了超参数的鲁棒性。
摘要：Parameter-efficient fine-tuning methods, such as Low-Rank Adaptation (LoRA), enable fast specialization of large pre-trained models to different downstream applications. However, this process often leads to catastrophic forgetting of the model's prior domain knowledge. We address this issue with LaLoRA, a weight-space regularization technique that applies a Laplace approximation to Low-Rank Adaptation. Our approach estimates the model's confidence in each parameter and constrains updates in high-curvature directions, preserving prior knowledge while enabling efficient target-domain learning. By applying the Laplace approximation only to the LoRA weights, the method remains lightweight. We evaluate LaLoRA by fine-tuning a Llama model for mathematical reasoning and demonstrate an improved learning-forgetting trade-off, which can be directly controlled via the method's regularization strength. We further explore different loss landscape curvature approximations for estimating parameter confidence, analyze the effect of the data used for the Laplace approximation, and study robustness across hyperparameters.
【3】Trust-Region Adaptive Policy Optimization
标题：信托区域适应性政策优化
链接：https://arxiv.org/abs/2512.17636
摘要：后训练方法，特别是监督微调（SFT）和强化学习（RL），在提高大型语言模型（LLM）的复杂推理能力方面发挥着重要作用。然而，占主导地位的两阶段流水线（SFT然后RL）存在一个关键的不一致性：SFT强制执行严格的模仿，抑制探索并导致遗忘，限制了RL的改进潜力。我们使用TRAPO（\textbf{T}rust-\textbf{R}egion \textbf{A}daptive \textbf{P}olicy \textbf{O}ptimization）解决了这种低效率问题，这是一种混合框架，通过优化专家前缀上的SFT损失和模型自身完成上的RL损失，在每个训练实例中交错SFT和RL，统一外部监督和自我探索。为了稳定训练，我们引入了信任区域SFT（TrSFT），它最大限度地减少了信任区域内的前向KL发散，但削弱了外部的优化，有效地向反向KL转移，并产生有利于RL的稳定的模式搜索更新。自适应前缀选择机制进一步基于测量的效用分配专家指导。五个数学推理基准的实验表明，TRAPO始终超过标准的SFT，RL和SFT-then-RL管道，以及最近的最先进的方法，为推理增强LLM建立了一个强大的新范式。
摘要：Post-training methods, especially Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL), play an important role in improving large language models' (LLMs) complex reasoning abilities. However, the dominant two-stage pipeline (SFT then RL) suffers from a key inconsistency: SFT enforces rigid imitation that suppresses exploration and induces forgetting, limiting RL's potential for improvements. We address this inefficiency with TRAPO (\textbf{T}rust-\textbf{R}egion \textbf{A}daptive \textbf{P}olicy \textbf{O}ptimization), a hybrid framework that interleaves SFT and RL within each training instance by optimizing SFT loss on expert prefixes and RL loss on the model's own completions, unifying external supervision and self-exploration. To stabilize training, we introduce Trust-Region SFT (TrSFT), which minimizes forward KL divergence inside a trust region but attenuates optimization outside, effectively shifting toward reverse KL and yielding stable, mode-seeking updates favorable for RL. An adaptive prefix-selection mechanism further allocates expert guidance based on measured utility. Experiments on five mathematical reasoning benchmarks show that TRAPO consistently surpasses standard SFT, RL, and SFT-then-RL pipelines, as well as recent state-of-the-art approaches, establishing a strong new paradigm for reasoning-enhanced LLMs.
【4】SHARP-QoS: Sparsely-gated Hierarchical Adaptive Routing for joint Prediction of QoS
标题：SHARP-Qos：用于联合预测Qos的稀疏选通分层自适应路由
链接：https://arxiv.org/abs/2512.17262
备注：12 pages, 4 figures, 10 tables
摘要 ：独立的面向服务的计算依赖于多个服务质量（QoS）参数，这些参数对于评估服务最优性至关重要。然而，现实世界的QoS数据是非常稀疏的，嘈杂的，并通过QoS的相互作用，地理和网络层面的因素所产生的层次依赖性，使准确的QoS预测的挑战。现有的方法通常单独预测每个QoS参数，需要多个相似的模型，这增加了计算成本，导致泛化能力差。虽然最近的联合QoS预测研究已经探索了共享的架构，但由于QoS参数之间的数值范围不一致所导致的损失缩放，并且进一步与不充分的表示学习作斗争，导致准确性降低，因此它们遭受负传递。本文提出了一种统一的联合QoS预测策略，称为SHARP-QoS，使用三个组件来解决这些问题。首先，我们引入了一种双重机制，通过庞加莱球中的双曲卷积从QoS和上下文结构中提取层次特征。其次，我们提出了一个自适应的功能共享机制，允许跨信息QoS和上下文信号的功能交换。一个门控特征融合模块，以支持结构和共享表示之间的动态特征选择。第三，我们设计了一个基于EMA的损失平衡策略，允许稳定的联合优化，从而减轻负转移。三个数据集上的两个，三个和四个QoS参数的评估表明，SHARP-QoS优于单任务和多任务基线。广泛的研究表明，我们的模型有效地解决了主要的挑战，包括稀疏性，鲁棒性离群值，冷启动，同时保持适度的计算开销，强调其可靠的联合QoS预测的能力。
摘要：Dependable service-oriented computing relies on multiple Quality of Service (QoS) parameters that are essential to assess service optimality. However, real-world QoS data are extremely sparse, noisy, and shaped by hierarchical dependencies arising from QoS interactions, and geographical and network-level factors, making accurate QoS prediction challenging. Existing methods often predict each QoS parameter separately, requiring multiple similar models, which increases computational cost and leads to poor generalization. Although recent joint QoS prediction studies have explored shared architectures, they suffer from negative transfer due to loss-scaling caused by inconsistent numerical ranges across QoS parameters and further struggle with inadequate representation learning, resulting in degraded accuracy. This paper presents an unified strategy for joint QoS prediction, called SHARP-QoS, that addresses these issues using three components. First, we introduce a dual mechanism to extract the hierarchical features from both QoS and contextual structures via hyperbolic convolution formulated in the Poincaré ball. Second, we propose an adaptive feature-sharing mechanism that allows feature exchange across informative QoS and contextual signals. A gated feature fusion module is employed to support dynamic feature selection among structural and shared representations. Third, we design an EMA-based loss balancing strategy that allows stable joint optimization, thereby mitigating the negative transfer. Evaluations on three datasets with two, three, and four QoS parameters demonstrate that SHARP-QoS outperforms both single- and multi-task baselines. Extensive study shows that our model effectively addresses major challenges, including sparsity, robustness to outliers, and cold-start, while maintaining moderate computational overhead, underscoring its capability for reliable joint QoS prediction.
【5】Learning to Plan, Planning to Learn: Adaptive Hierarchical RL-MPC for Sample-Efficient Decision Making
标题：学会计划，计划学习：自适应分层WL-MPC，用于样本高效决策
链接：https://arxiv.org/abs/2512.17091
备注：23 pages, 8 figures. Under review
摘要：我们提出了一种新的方法来解决规划问题的层次结构，融合强化学习和MPC规划。我们的公式紧密而优雅地耦合了两个规划范式。它利用强化学习动作来通知MPPI采样器，并自适应地聚合MPPI样本来通知值估计。由此产生的自适应过程利用了进一步的MPPI探索，其中值估计是不确定的，并提高了训练的鲁棒性和整体结果的政策。这导致一个强大的规划方法，可以处理复杂的规划问题，并很容易地适应不同的应用程序，如在几个领域，包括赛车，修改的Acrobot和月球着陆器与添加的障碍。我们在这些领域的研究结果显示，在奖励和任务成功方面，数据效率和整体性能都更好，与现有方法相比，成功率提高了72%，与非自适应采样相比，收敛速度加快（x2.1）。
摘要：We propose a new approach for solving planning problems with a hierarchical structure, fusing reinforcement learning and MPC planning. Our formulation tightly and elegantly couples the two planning paradigms. It leverages reinforcement learning actions to inform the MPPI sampler, and adaptively aggregates MPPI samples to inform the value estimation. The resulting adaptive process leverages further MPPI exploration where value estimates are uncertain, and improves training robustness and the overall resulting policies. This results in a robust planning approach that can handle complex planning problems and easily adapts to different applications, as demonstrated over several domains, including race driving, modified Acrobot, and Lunar Lander with added obstacles. Our results in these domains show better data efficiency and overall performance in terms of both rewards and task success, with up to a 72% increase in success rate compared to existing approaches, as well as accelerated convergence (x2.1) compared to non-adaptive sampling.
【6】Bandwidth-Efficient Adaptive Mixture-of-Experts via Low-Rank Compensation
标题：通过低等级补偿实现带宽高效的自适应专家混合
链接：https://arxiv.org/abs/2512.17073
摘要：混合专家（MoE）模型通过稀疏激活但压力记忆和带宽来扩展容量。卸载通过按需获取专家来减少GPU内存，但令牌级路由会导致不规则的传输，从而使推理I/O受限。静态均匀量化通过忽略专家异质性减少了流量，但在积极压缩下降低了准确性。我们提出了带宽有效的自适应混合专家通过低秩补偿，它使用预先计算的低秩补偿器进行路由器引导的精度恢复。在推理时，我们的方法将紧凑的低秩因子与每个令牌的Top-n（n
摘要：Mixture-of-Experts (MoE) models scale capacity via sparse activation but stress memory and bandwidth. Offloading alleviates GPU memory by fetching experts on demand, yet token-level routing causes irregular transfers that make inference I/O-bound. Static uniform quantization reduces traffic but degrades accuracy under aggressive compression by ignoring expert heterogeneity. We present Bandwidth-Efficient Adaptive Mixture-of-Experts via Low-Rank Compensation, which performs router-guided precision restoration using precomputed low-rank compensators. At inference time, our method transfers compact low-rank factors with Top-n (n
强化学习(3篇)
【1】Assessing Long-Term Electricity Market Design for Ambitious Decarbonization Targets using Multi-Agent Reinforcement Learning
标题：使用多智能体强化学习评估雄心勃勃的脱碳目标的长期电力市场设计
链接：https://arxiv.org/abs/2512.17444
备注：Accepted to Energy and AI. Code available in https://github.com/jjgonzalez2491/MARLEY_V1
摘要：电力系统是将当今社会转变为无碳经济的关键。长期的电力市场机制，包括拍卖、支持计划和其他政策工具，对形成发电组合至关重要。鉴于需要更先进的工具来支持政策制定者和其他利益相关者设计，测试和评估长期市场，这项工作提出了一个多智能体强化学习模型，能够捕捉脱碳能源系统的关键特征。利润最大化的发电公司在批发电力市场上做出投资决策，以响应系统需求，竞争动态和政策信号。该模型采用独立的最近端策略优化，这是选择适合分散和竞争环境。尽管如此，考虑到多智能体环境中独立学习的固有挑战，广泛的超参数搜索可以确保分散式训练产生与竞争行为一致的市场结果。该模型适用于意大利电力系统的程式化版本，并在不同水平的竞争，市场设计和政策方案下进行测试。结果突出了市场设计对电力部门脱碳和避免价格波动的关键作用。拟议的框架允许评估长期电力市场，其中多种政策和市场机制同时相互作用，市场参与者响应和适应脱碳途径。
摘要 ：Electricity systems are key to transforming today's society into a carbon-free economy. Long-term electricity market mechanisms, including auctions, support schemes, and other policy instruments, are critical in shaping the electricity generation mix. In light of the need for more advanced tools to support policymakers and other stakeholders in designing, testing, and evaluating long-term markets, this work presents a multi-agent reinforcement learning model capable of capturing the key features of decarbonizing energy systems. Profit-maximizing generation companies make investment decisions in the wholesale electricity market, responding to system needs, competitive dynamics, and policy signals. The model employs independent proximal policy optimization, which was selected for suitability to the decentralized and competitive environment. Nevertheless, given the inherent challenges of independent learning in multi-agent settings, an extensive hyperparameter search ensures that decentralized training yields market outcomes consistent with competitive behavior. The model is applied to a stylized version of the Italian electricity system and tested under varying levels of competition, market designs, and policy scenarios. Results highlight the critical role of market design for decarbonizing the electricity sector and avoiding price volatility. The proposed framework allows assessing long-term electricity markets in which multiple policy and market mechanisms interact simultaneously, with market participants responding and adapting to decarbonization pathways.
【2】GB-DQN: Gradient Boosted DQN Models for Non-stationary Reinforcement Learning
标题：GB-DQN：用于非平稳强化学习的梯度增强DQN模型
链接：https://arxiv.org/abs/2512.17034
备注：23 pages. Submitted to Machine Learning
摘要：非平稳环境对深度强化学习提出了根本性的挑战，因为动态或奖励的变化会使学习到的价值函数失效，并导致灾难性的遗忘。我们提出了一种自适应集成方法，通过增量残差学习来解决模型漂移问题。GB-DQN不是对单个Q网络进行再训练，而是构建一个加性集成，其中每个新的学习者都被训练为近似漂移后当前集成的Bellman残差。我们提供的理论结果表明，每个提升步骤减少了经验贝尔曼残差和合奏收敛到漂移后的最佳值函数在标准假设下。与DQN和常见的非平稳基线相比，具有受控动态变化的各种控制任务的实验表明了更快的恢复，更好的稳定性和更强的鲁棒性。
摘要：Non-stationary environments pose a fundamental challenge for deep reinforcement learning, as changes in dynamics or rewards invalidate learned value functions and cause catastrophic forgetting. We propose \emph{Gradient-Boosted Deep Q-Networks (GB-DQN)}, an adaptive ensemble method that addresses model drift through incremental residual learning. Instead of retraining a single Q-network, GB-DQN constructs an additive ensemble in which each new learner is trained to approximate the Bellman residual of the current ensemble after drift. We provide theoretical results showing that each boosting step reduces the empirical Bellman residual and that the ensemble converges to the post-drift optimal value function under standard assumptions. Experiments across a diverse set of control tasks with controlled dynamics changes demonstrate faster recovery, improved stability, and greater robustness compared to DQN and common non-stationary baselines.
【3】HydroGym: A Reinforcement Learning Platform for Fluid Dynamics
标题：HydroGym：流体动力学强化学习平台
链接：https://arxiv.org/abs/2512.17534
摘要：流体流动的建模和控制对于包括交通、能源和医学在内的多个科学和工程领域至关重要。有效的流量控制可以导致，例如，升力增加、阻力减小、混合增强和噪声减小。然而，控制流体面临着几个重大挑战，包括空间和时间中的高维，非线性和多尺度相互作用。强化学习（RL）最近在机器人和蛋白质折叠等复杂领域取得了巨大成功，但由于缺乏标准化的基准平台和流体模拟的计算需求，其在流动控制中的应用受到阻碍。为了应对这些挑战，我们引入了HydroGym，这是一个用于流量控制研究的独立于求解器的RL平台。HydroGym集成了复杂的流控制基准，可扩展的运行时基础设施和最先进的RL算法。我们的平台包括42个经过验证的环境，从典型的层流到复杂的三维湍流场景，在广泛的雷诺数范围内进行验证。我们为传统RL和可微求解器提供不可微求解器，通过梯度增强优化显著提高样本效率。综合评估表明，RL代理一致地发现跨配置的鲁棒控制原则，如边界层操纵，声反馈中断和尾流重组。迁移学习研究表明，在一个雷诺数或几何形状下学习的控制器可以有效地适应新的条件，所需的训练次数减少约50%。HydroGym平台具有高度的可扩展性和可扩展性，为流体动力学、机器学习和控制领域的研究人员提供了一个框架，以添加环境、代理模型和控制算法，从而推动科学技术的发展。
摘要：Modeling and controlling fluid flows is critical for several fields of science and engineering, including transportation, energy, and medicine. Effective flow control can lead to, e.g., lift increase, drag reduction, mixing enhancement, and noise reduction. However, controlling a fluid faces several significant challenges, including high-dimensional, nonlinear, and multiscale interactions in space and time. Reinforcement learning (RL) has recently shown great success in complex domains, such as robotics and protein folding, but its application to flow control is hindered by a lack of standardized benchmark platforms and the computational demands of fluid simulations. To address these challenges, we introduce HydroGym, a solver-independent RL platform for flow control research. HydroGym integrates sophisticated flow control benchmarks, scalable runtime infrastructure, and state-of-the-art RL algorithms. Our platform includes 42 validated environments spanning from canonical laminar flows to complex three-dimensional turbulent scenarios, validated over a wide range of Reynolds numbers. We provide non-differentiable solvers for traditional RL and differentiable solvers that dramatically improve sample efficiency through gradient-enhanced optimization. Comprehensive evaluation reveals that RL agents consistently discover robust control principles across configurations, such as boundary layer manipulation, acoustic feedback disruption, and wake reorganization. Transfer learning studies demonstrate that controllers learned at one Reynolds number or geometry adapt efficiently to new conditions, requiring approximately 50% fewer training episodes. The HydroGym platform is highly extensible and scalable, providing a framework for researchers in fluid dynamics, machine learning, and control to add environments, surrogate models, and control algorithms to advance science and technology.
医学相关(9篇)
【1】When De-noising Hurts: A Systematic Study of Speech Enhancement Effects on Modern Medical ASR Systems
标题：去噪何时会造成伤害：现代医疗ASB系统语音增强效果的系统研究
链接：https://arxiv.org/abs/2512.17562
备注：Technical Report
摘要：语音增强方法被普遍认为可以提高噪声环境中自动语音识别（ASR）的性能。然而，对于在多样化、有噪数据上训练的现代大规模ASR模型来说，这些技术的有效性不能被视为理所当然。我们在四个最先进的ASR系统上对MetricGAN加语音库去噪进行了系统评估：OpenAI Whisper，NVIDIA Parakeet，Google Gemini Flash 2.0，Parrotlet-a，在九种噪声条件下使用500个医疗语音记录。ASR性能是使用语义WER（semWER）来衡量的，这是一个归一化的单词错误率（WER）指标，用于解释特定于域的标准化。我们的研究结果揭示了一个违反直觉的发现：语音增强预处理降低了所有噪声条件和模型的ASR性能。在所有40种测试配置（4种型号x10种条件）中，原始嘈杂音频的semWER低于增强音频，衰减范围为1.1%至46.6%的绝对semWER增加。这些研究结果表明，现代ASR模型具有足够的内部噪声鲁棒性，传统的语音增强可能会删除声学特征的关键ASR。对于在嘈杂的临床环境中部署医疗抄写系统的从业者来说，我们的研究结果表明，使用降噪技术对音频进行预处理不仅可能在计算上浪费，而且可能对转录准确性有害。
摘要 ：Speech enhancement methods are commonly believed to improve the performance of automatic speech recognition (ASR) in noisy environments. However, the effectiveness of these techniques cannot be taken for granted in the case of modern large-scale ASR models trained on diverse, noisy data. We present a systematic evaluation of MetricGAN-plus-voicebank denoising on four state-of-the-art ASR systems: OpenAI Whisper, NVIDIA Parakeet, Google Gemini Flash 2.0, Parrotlet-a using 500 medical speech recordings under nine noise conditions. ASR performance is measured using semantic WER (semWER), a normalized word error rate (WER) metric accounting for domain-specific normalizations. Our results reveal a counterintuitive finding: speech enhancement preprocessing degrades ASR performance across all noise conditions and models. Original noisy audio achieves lower semWER than enhanced audio in all 40 tested configurations (4 models x 10 conditions), with degradations ranging from 1.1% to 46.6% absolute semWER increase. These findings suggest that modern ASR models possess sufficient internal noise robustness and that traditional speech enhancement may remove acoustic features critical for ASR. For practitioners deploying medical scribe systems in noisy clinical environments, our results indicate that preprocessing audio with noise reduction techniques might not just be computationally wasteful but also be potentially harmful to the transcription accuracy.
【2】PathBench-MIL: A Comprehensive AutoML and Benchmarking Framework for Multiple Instance Learning in Histopathology
标题：PathBench-MIL：用于组织病理学多实例学习的全面AutoML和基准框架
链接：https://arxiv.org/abs/2512.17517
备注：14 Pages, 3 Figures, 2 Appendices
摘要：我们介绍PathBench-MIL，这是一个开源的AutoML和基准框架，用于组织病理学中的多实例学习（MIL）。该系统自动化端到端MIL管道构建，包括预处理，特征提取和MIL聚合，并提供数十个MIL模型和特征提取器的可重复基准测试。PathBench-MIL集成了可视化工具、统一的配置系统和模块化可扩展性，可跨数据集和任务进行快速实验和标准化。PathBench-MIL可在https://github.com/Sbrussee/PathBench-MIL上公开获取
摘要：We introduce PathBench-MIL, an open-source AutoML and benchmarking framework for multiple instance learning (MIL) in histopathology. The system automates end-to-end MIL pipeline construction, including preprocessing, feature extraction, and MIL-aggregation, and provides reproducible benchmarking of dozens of MIL models and feature extractors. PathBench-MIL integrates visualization tooling, a unified configuration system, and modular extensibility, enabling rapid experimentation and standardization across datasets and tasks. PathBench-MIL is publicly available at https://github.com/Sbrussee/PathBench-MIL
【3】TwinSegNet: A Digital Twin-Enabled Federated Learning Framework for Brain Tumor Analysis
标题：TwinSegNet：用于脑肿瘤分析的数字双启用联邦学习框架
链接：https://arxiv.org/abs/2512.17488
备注：IEEE Virtual Conference on Communications. 4-6 November 2025
摘要：脑肿瘤分割在疾病的诊断和治疗计划中至关重要。然而，目前的深度学习方法依赖于集中的数据收集，这引起了隐私问题，并限制了不同机构的泛化。在本文中，我们提出了TwinSegNet，这是一个隐私保护的联邦学习框架，它将混合ViT-UNet模型与个性化的数字双胞胎集成在一起，以实现准确和实时的脑肿瘤分割。我们的架构将卷积编码器与Vision Transformer瓶颈相结合，以捕获本地和全局上下文。每个机构都对私人数据的全球模型进行微调，以形成其数字孪生模型。在九个异构MRI数据集上进行评估，包括BraTS 2019-2021和自定义肿瘤集合，TwinSegNet实现了高Dice评分（高达0.90%）和超过90%的灵敏度/特异性，证明了非独立同分布（IID）客户端分布的鲁棒性。与TumorVisNet等集中式模型的比较结果突出了TwinSegNet在保护隐私而不牺牲性能方面的有效性。我们的方法可以为多机构临床环境提供可扩展的个性化细分，同时遵守严格的数据保密要求。
摘要：Brain tumor segmentation is critical in diagnosis and treatment planning for the disease. Yet, current deep learning methods rely on centralized data collection, which raises privacy concerns and limits generalization across diverse institutions. In this paper, we propose TwinSegNet, which is a privacy-preserving federated learning framework that integrates a hybrid ViT-UNet model with personalized digital twins for accurate and real-time brain tumor segmentation. Our architecture combines convolutional encoders with Vision Transformer bottlenecks to capture local and global context. Each institution fine-tunes the global model of private data to form its digital twin. Evaluated on nine heterogeneous MRI datasets, including BraTS 2019-2021 and custom tumor collections, TwinSegNet achieves high Dice scores (up to 0.90%) and sensitivity/specificity exceeding 90%, demonstrating robustness across non-independent and identically distributed (IID) client distributions. Comparative results against centralized models such as TumorVisNet highlight TwinSegNet's effectiveness in preserving privacy without sacrificing performance. Our approach enables scalable, personalized segmentation for multi-institutional clinical settings while adhering to strict data confidentiality requirements.
【4】Alzheimer's Disease Brain Network Mining
标题：阿尔茨海默病脑网络挖掘
链接：https://arxiv.org/abs/2512.17276
摘要：用于阿尔茨海默病（AD）诊断的机器学习方法面临着根本性的挑战。临床评估是昂贵的和侵入性的，只留下一小部分神经成像数据集的真实标签。我们介绍了用于异构阿尔茨海默病的多视图自适应传输聚类（MATCH-AD），这是一个半监督框架，它集成了深度表示学习，基于图的标签传播和最佳传输理论来解决这一限制。该框架利用神经成像数据中的流形结构将诊断信息从有限的标记样本传播到更大的未标记人群，同时使用Wasserstein距离来量化认知状态之间的疾病进展。对来自国家阿尔茨海默氏症协调中心的近5000名受试者进行了评估，包括来自数百个大脑区域的结构MRI测量，脑脊液生物标志物和临床变量MATCHAD实现了近乎完美的诊断准确性，尽管只有不到三分之一的受试者有地面真实标签。该框架大大优于所有的基线方法，实现Kappa表示几乎完美的协议相比，弱协议的最佳基线，诊断可靠性的定性转变。即使在严重的标签稀缺的临床上仍然有用的性能，我们提供了理论上的收敛保证，证明标签传播误差和传输稳定性的界限。这些结果表明，有原则的半监督学习可以释放全球积累的部分注释的神经成像数据的巨大存储库的诊断潜力，大大减少注释负担，同时保持适合临床部署的准确性。
摘要：Machine learning approaches for Alzheimer's disease (AD) diagnosis face a fundamental challenges. Clinical assessments are expensive and invasive, leaving ground truth labels available for only a fraction of neuroimaging datasets. We introduce Multi view Adaptive Transport Clustering for Heterogeneous Alzheimer's Disease (MATCH-AD), a semi supervised framework that integrates deep representation learning, graph-based label propagation, and optimal transport theory to address this limitation. The framework leverages manifold structure in neuroimaging data to propagate diagnostic information from limited labeled samples to larger unlabeled populations, while using Wasserstein distances to quantify disease progression between cognitive states. Evaluated on nearly five thousand subjects from the National Alzheimer's Coordinating Center, encompassing structural MRI measurements from hundreds of brain regions, cerebrospinal fluid biomarkers, and clinical variables MATCHAD achieves near-perfect diagnostic accuracy despite ground truth labels for less than one-third of subjects. The framework substantially outperforms all baseline methods, achieving kappa indicating almost perfect agreement compared to weak agreement for the best baseline, a qualitative transformation in diagnostic reliability. Performance remains clinically useful even under severe label scarcity, and we provide theoretical convergence guarantees with proven bounds on label propagation error and transport stability. These results demonstrate that principled semi-supervised learning can unlock the diagnostic potential of the vast repositories of partially annotated neuroimaging data accumulating worldwide, substantially reducing annotation burden while maintaining accuracy suitable for clinical deployment.
【5】The Effect of Negation on CLIP in Medical Imaging: Limitations of Contrastive Language-Image Pretraining
标题：医学成像中否定对CLIP的影响：对比图像预训练的局限性
链接：https://arxiv.org/abs/2512.17121
备注：10 pages, 7 figures, submitted to WACV Pixels to Patients Workshop
摘要：像CLIP这样的大型视觉语言模型越来越多地用于医学成像任务，因为它们能够在不需要大量标记数据的情况下对齐图像和文本。这使得它们特别适用于临床环境中的图像检索、报告生成和分类等应用。这种方法的一个潜在问题是，基于CLIP的模型在解释否定短语时往往表现不佳，这在医学诊断的背景下尤其成问题。在这项研究中，我们评估了斯坦福AIMI CheXagent模型使用有否定和无否定提示正确检索胸部X射线图像的能力。该项目的目标是了解该模型失败的地方，然后将其用作基础模型，通过微调以前工作中概述的方法来提高其检索精度。本研究的结果表明，CLIP模型中的否定处理有所改善，但积极提示评估的准确性略有下降。除了检索准确性外，我们还通过标记归因，t-SNE投影和注意力头消融来检查内部模型行为，以更好地表征每种微调方法如何重塑否定临床语言的文本编码器表示。通过这项工作，我们希望更好地了解CLIP的内部行为，并使用临床相关语言改进其否定处理，以提高其在医疗AI设备中的可靠性。
摘要：Large vision-language models like CLIP are increasingly used in medical imaging tasks due to their ability to align images and text without the need for extensive labeled data. This makes them particularly useful for applications like image retrieval, report generation, and classification in clinical settings. A potential issue to this approach is that CLIP-based models often under perform when interpreting negated phrases, which is especially problematic in the context of medical diagnosing. In this study, we evaluate the Stanford AIMI CheXagent model on its ability to correctly retrieve chest X-ray images using prompts with and without negation. The goal of this project is to understand where this model fails and then use it as a base model to improve its retrieval accuracy by fine tuning methods outlined in previous work. Results from this study show improvement in handling of negation in the CLIP model with a slight decrease in accuracy of positive prompt evaluation. Alongside retrieval accuracy, we examined internal model behavior through token attribution, t-SNE projection, and attention-head ablation to better characterize how each fine tuning approach reshaped the text encoders representation of negated clinical language. Through this work, we hope to better understand the internal behavior of CLIP and improve its handling of negation using clinically relevant language for improving its reliability in medical AI devices.
【6】UniCoMTE: A Universal Counterfactual Framework for Explaining Time-Series Classifiers on ECG Data
标题：UniCoMTE：用于解释心电图数据时间序列分类器的通用反事实框架
链接：https://arxiv.org/abs/2512.17100
备注：21 pages, 7 figures
摘要：机器学习模型，特别是深度神经网络，在对复杂时间序列数据进行分类方面表现出了强大的性能。然而，它们的暗箱性质限制了信任和采用，特别是在医疗保健等高风险领域。为了应对这一挑战，我们引入了UniCoMTE，这是一个模型不可知的框架，用于为多变量时间序列分类器生成反事实解释。该框架通过修改输入样本并评估其对模型预测的影响来确定对模型预测影响最大的时间特征。UniCoMTE与各种模型架构兼容，并直接对原始时间序列输入进行操作。在这项研究中，我们评估UniCoMTE的解释的时间序列心电图分类。我们量化的解释质量，通过比较我们的解释的可理解性，可理解性的既定技术（石灰和SHAP），并评估其推广到类似的样本。此外，通过由医学专家完成的问卷评估临床效用，这些医学专家审查与原始ECG样本一起提供的反事实解释。结果表明，我们的方法产生了简洁，稳定和人性化的解释，在清晰度和适用性方面优于现有的方法。通过将模型预测与有意义的信号模式联系起来，该框架提高了深度学习模型在现实世界时间序列应用中的可解释性。
摘要：Machine learning models, particularly deep neural networks, have demonstrated strong performance in classifying complex time series data. However, their black-box nature limits trust and adoption, especially in high-stakes domains such as healthcare. To address this challenge, we introduce UniCoMTE, a model-agnostic framework for generating counterfactual explanations for multivariate time series classifiers. The framework identifies temporal features that most heavily influence a model's prediction by modifying the input sample and assessing its impact on the model's prediction. UniCoMTE is compatible with a wide range of model architectures and operates directly on raw time series inputs. In this study, we evaluate UniCoMTE's explanations on a time series ECG classifier. We quantify explanation quality by comparing our explanations' comprehensibility to comprehensibility of established techniques (LIME and SHAP) and assessing their generalizability to similar samples. Furthermore, clinical utility is assessed through a questionnaire completed by medical experts who review counterfactual explanations presented alongside original ECG samples. Results show that our approach produces concise, stable, and human-aligned explanations that outperform existing methods in both clarity and applicability. By linking model predictions to meaningful signal patterns, the framework advances the interpretability of deep learning models for real-world time series applications.
【7】Breast Cancer Neoadjuvant Chemotherapy Treatment Response Prediction Using Aligned Longitudinal MRI and Clinical Data
标题：使用对齐纵向MRI和临床数据预测乳腺癌新辅助化疗治疗反应
链接：https://arxiv.org/abs/2512.17759
摘要：目的：本研究采用纵向对比增强磁共振成像（CE-MRI）和临床数据，研究乳腺癌患者对新辅助化疗（NACT）的治疗反应预测。目标是开发机器学习（ML）模型来预测病理完全缓解（PCR二元分类）和5年无复发生存状态（RFS二元分类）。方法：提出的框架包括肿瘤分割，图像配准，特征提取和预测建模。使用图像配准方法，可以从原始肿瘤部位在不同时间点提取MRI图像特征并进行比较，从而监测NACT过程中的肿瘤内变化。实现并比较了四种特征提取器，包括一种放射组学和三种基于深度学习的特征提取器（MedicalNet，Segformer 3D，SAM-Med 3D）。结合三种特征选择方法和四种ML模型，建立预测模型并进行比较。结果：所提出的基于图像配准的特征提取一致地改善了预测模型。在PCR和RFS分类任务中，在放射组学特征上训练的逻辑回归模型表现最好，PCR分类的AUC为0.88，分类准确度为0.85，RFS分类的AUC为0.78，分类准确度为0.72。结论：结果表明，图像配准方法在预测PCR和RFS的纵向特征学习中具有显着改善的性能。放射组学特征提取器比预训练的深度学习特征提取器更有效，具有更高的性能和更好的可解释性。
摘要：Aim: This study investigates treatment response prediction to neoadjuvant chemotherapy (NACT) in breast cancer patients, using longitudinal contrast-enhanced magnetic resonance images (CE-MRI) and clinical data. The goal is to develop machine learning (ML) models to predict pathologic complete response (PCR binary classification) and 5-year relapse-free survival status (RFS binary classification). Method: The proposed framework includes tumour segmentation, image registration, feature extraction, and predictive modelling. Using the image registration method, MRI image features can be extracted and compared from the original tumour site at different time points, therefore monitoring the intratumor changes during NACT process. Four feature extractors, including one radiomics and three deep learning-based (MedicalNet, Segformer3D, SAM-Med3D) were implemented and compared. In combination with three feature selection methods and four ML models, predictive models are built and compared. Results: The proposed image registration-based feature extraction consistently improves the predictive models. In the PCR and RFS classification tasks logistic regression model trained on radiomic features performed the best with an AUC of 0.88 and classification accuracy of 0.85 for PCR classification, and AUC of 0.78 and classification accuracy of 0.72 for RFS classification. Conclusions: It is evidenced that the image registration method has significantly improved performance in longitudinal feature learning in predicting PCR and RFS. The radiomics feature extractor is more effective than the pre-trained deep learning feature extractors, with higher performance and better interpretability.
【8】Resource-efficient medical image classification for edge devices
标题：边缘设备资源高效的医学图像分类
链接：https://arxiv.org/abs/2512.17515
备注：Conference paper published in ICAMIDA 2025 (IEEE)
摘要：医学图像分类是医疗保健中的一项关键任务，可以实现准确和及时的诊断。然而，由于计算和内存的限制，在资源受限的边缘设备上部署深度学习模型会带来重大挑战。本研究探讨一种资源有效的方法，医学影像分类，采用模型量化技术。量化降低了模型参数和激活的精度，在不牺牲分类精度的情况下显著降低了计算开销和内存需求。该研究重点关注为边缘设备量身定制的量化感知训练（QAT）和训练后量化（PTQ）方法的优化，分析它们对医学成像数据集模型性能的影响。实验结果表明，量化模型实现了模型大小和推理延迟的大幅减少，从而实现了边缘硬件上的实时处理，同时保持了临床上可接受的诊断准确性。这项工作为在远程和资源有限的环境中部署人工智能驱动的医疗诊断提供了一条实用的途径，提高了医疗保健技术的可访问性和可扩展性。
摘要：Medical image classification is a critical task in healthcare, enabling accurate and timely diagnosis. However, deploying deep learning models on resource-constrained edge devices presents significant challenges due to computational and memory limitations. This research investigates a resource-efficient approach to medical image classification by employing model quantization techniques. Quantization reduces the precision of model parameters and activations, significantly lowering computational overhead and memory requirements without sacrificing classification accuracy. The study focuses on the optimization of quantization-aware training (QAT) and post-training quantization (PTQ) methods tailored for edge devices, analyzing their impact on model performance across medical imaging datasets. Experimental results demonstrate that quantized models achieve substantial reductions in model size and inference latency, enabling real-time processing on edge hardware while maintaining clinically acceptable diagnostic accuracy. This work provides a practical pathway for deploying AI-driven medical diagnostics in remote and resource-limited settings, enhancing the accessibility and scalability of healthcare technologies.
【9】Penalized Fair Regression for Multiple Groups in Chronic Kidney Disease
标题：慢性肾病多个群体的公平回归受到惩罚
链接：https://arxiv.org/abs/2512.17340
摘要：公平回归方法有可能减轻医疗保健中的社会偏见问题，但当多个群体经历这种偏见时，惩罚公平回归的工作很少。我们提出了一个通用的回归框架，解决了这一差距与不公平的惩罚多组。我们的方法证明了二元结果与真正的积极率差距的处罚。它可以有效地实施，通过减少成本敏感的分类问题。我们还引入了新的评分函数，用于自动选择惩罚权重。我们的惩罚公平回归方法在模拟中进行了实证研究，在那里它们实现了超越现有比较方法的公平性-准确性边界。最后，我们将这些方法应用于一项全国性的慢性肾脏病多中心初级保健研究，以开发一个公平的终末期肾脏病分类器。在那里，我们发现，在公平性方面有实质性的改善，多个种族和民族群体在医疗保健系统中经历社会偏见，而没有任何明显的损失，在整体上适合。
摘要：Fair regression methods have the potential to mitigate societal bias concerns in health care, but there has been little work on penalized fair regression when multiple groups experience such bias. We propose a general regression framework that addresses this gap with unfairness penalties for multiple groups. Our approach is demonstrated for binary outcomes with true positive rate disparity penalties. It can be efficiently implemented through reduction to a cost-sensitive classification problem. We additionally introduce novel score functions for automatically selecting penalty weights. Our penalized fair regression methods are empirically studied in simulations, where they achieve a fairness-accuracy frontier beyond that of existing comparison methods. Finally, we apply these methods to a national multi-site primary care study of chronic kidney disease to develop a fair classifier for end-stage renal disease. There we find substantial improvements in fairness for multiple race and ethnicity groups who experience societal bias in the health care system without any appreciable loss in overall fit.
推荐(2篇)
【1】Exploiting ID-Text Complementarity via Ensembling for Sequential Recommendation
标题：通过组合利用ID文本互补性进行顺序推荐
链接：https://arxiv.org/abs/2512.17820
摘要：现代顺序推荐（SR）模型通常利用模态特征来表示项目，这在很大程度上是由语言和视觉建模的最新进展所激发的。为此，一些作品用模态嵌入完全取代了ID嵌入，声称模态嵌入使ID嵌入变得不必要，因为它们可以匹配甚至超过ID嵌入性能。另一方面，许多作品联合利用ID和模态特征，但认为复杂的融合策略，如多阶段训练和/或复杂的对齐架构，是必要的，这种联合利用。然而，这两个工作线的基础是缺乏对ID和模态特征的互补性的理解。在这项工作中，我们通过研究ID和基于文本的SR模型的互补性来解决这一差距。我们证明了这些模型确实可以学习互补信号，这意味着当与另一个一起正确使用时，任何一个都应该提供性能增益。出于这一动机，我们提出了一种新的SR方法，通过独立的模型训练保持ID文本的互补性，然后通过一个简单的集成策略利用它。尽管这种方法的简单性，我们表明它优于几个有竞争力的SR基线，这意味着ID和文本功能是必要的，以实现最先进的SR性能，但复杂的融合架构不是。
摘要：Modern Sequential Recommendation (SR) models commonly utilize modality features to represent items, motivated in large part by recent advancements in language and vision modeling. To do so, several works completely replace ID embeddings with modality embeddings, claiming that modality embeddings render ID embeddings unnecessary because they can match or even exceed ID embedding performance. On the other hand, many works jointly utilize ID and modality features, but posit that complex fusion strategies, such as multi-stage training and/or intricate alignment architectures, are necessary for this joint utilization. However, underlying both these lines of work is a lack of understanding of the complementarity of ID and modality features. In this work, we address this gap by studying the complementarity of ID- and text-based SR models. We show that these models do learn complementary signals, meaning that either should provide performance gain when used properly alongside the other. Motivated by this, we propose a new SR method that preserves ID-text complementarity through independent model training, then harnesses it through a simple ensembling strategy. Despite this method's simplicity, we show it outperforms several competitive SR baselines, implying that both ID and text features are necessary to achieve state-of-the-art SR performance but complex fusion architectures are not.
【2】Warmer for Less: A Cost-Efficient Strategy for Cold-Start Recommendations at Pinterest
标题：以更少的价格温暖：Pinterest冷启动推荐的经济高效策略
链接：https://arxiv.org/abs/2512.17277
备注：Submitted to the WWW'26
摘要：Pinterest是一个领先的视觉发现平台，推荐系统（RecSys）是向我们的用户提供相关，引人入胜和新鲜内容的关键。在本文中，我们研究了改进RecSys模型预测冷启动（CS）项目的问题，这些项目在训练数据中很少出现。虽然这个问题在学术界得到了很好的研究，但很少有研究能在Pinterest这样的平台上有效地解决其根本原因。通过调查实时交通数据，我们确定了CS问题的几个挑战，并为每个挑战开发了相应的解决方案：首先，工业规模的RecSys模型必须在严格的计算约束下运行。由于CS项目是少数，任何相关的改进都必须具有很高的成本效益。为了解决这个问题，我们的解决方案被设计成轻量级的，总参数仅增加了5%。第二，CS项仅由非历史（例如，内容或属性）特征，模型通常将其视为不太重要。为了提升它们的重要性，我们为非历史特征引入了一个剩余连接。第三，与非CS项目相比，CS项目倾向于获得较低的预测分数，从而降低了它们出现的可能性。我们通过将分数正则化项纳入模型来缓解这一点。第四，与CS项目相关的标签是稀疏的，使得模型很难从中学习。我们应用流形混合技术来解决这种数据稀疏性。通过共同实施，我们的方法将Pinterest的新鲜内容参与度提高了10%，而不会对整体参与度和成本产生负面影响，并已部署到Pinterest上为超过5.7亿用户提供服务。
摘要 ：Pinterest is a leading visual discovery platform where recommender systems (RecSys) are key to delivering relevant, engaging, and fresh content to our users. In this paper, we study the problem of improving RecSys model predictions for cold-start (CS) items, which appear infrequently in the training data. Although this problem is well-studied in academia, few studies have addressed its root causes effectively at the scale of a platform like Pinterest. By investigating live traffic data, we identified several challenges of the CS problem and developed a corresponding solution for each: First, industrial-scale RecSys models must operate under tight computational constraints. Since CS items are a minority, any related improvements must be highly cost-efficient. To address this, our solutions were designed to be lightweight, collectively increasing the total parameters by only 5%. Second, CS items are represented only by non-historical (e.g., content or attribute) features, which models often treat as less important. To elevate their significance, we introduce a residual connection for the non-historical features. Third, CS items tend to receive lower prediction scores compared to non-CS items, reducing their likelihood of being surfaced. We mitigate this by incorporating a score regularization term into the model. Fourth, the labels associated with CS items are sparse, making it difficult for the model to learn from them. We apply the manifold mixup technique to address this data sparsity. Implemented together, our methods increased fresh content engagement at Pinterest by 10% without negatively impacting overall engagement and cost, and have been deployed to serve over 570 million users on Pinterest.
聚类(3篇)
【1】MAD-OOD: A Deep Learning Cluster-Driven Framework for an Out-of-Distribution Malware Detection and Classification
标题：MAD-OOD：一个深度学习机器人驱动的框架，用于分发外恶意软件检测和分类
链接：https://arxiv.org/abs/2512.17594
摘要：由于多态和变形恶意软件变体引入了大量的家族内变异性，分发外（OOD）检测仍然是恶意软件分类中的一个关键挑战。大多数现有的基于深度学习的恶意软件检测器依赖于封闭世界假设，无法充分建模这种类内变化，导致在面对以前看不见的恶意软件家族时性能下降。本文介绍了MADOOD，一种新的两阶段，集群驱动的深度学习框架，用于强大的OOD恶意软件检测和分类。在第一阶段中，恶意软件家族嵌入使用从高斯判别分析（GDA）导出的类条件球形决策边界来建模，从而在训练期间不需要OOD数据的情况下实现分布和OOD样本的统计接地分离。采用基于Z得分的多类质心距离分析来可靠地识别潜在空间中的异常样本。在第二阶段，深度神经网络集成了基于聚类的预测、精细嵌入和监督分类器输出，以提高最终分类精度。对包括25个已知家族和多个新型OOD变体的基准恶意软件数据集进行的广泛评估表明，MADOOD显著优于最先进的OOD检测方法，在看不见的恶意软件家族上实现了高达0.911的AUC。所提出的框架为不断发展的网络安全环境中的真实世界恶意软件检测和异常识别提供了可扩展的，可解释的和统计原则的解决方案。
摘要：Out of distribution (OOD) detection remains a critical challenge in malware classification due to the substantial intra family variability introduced by polymorphic and metamorphic malware variants. Most existing deep learning based malware detectors rely on closed world assumptions and fail to adequately model this intra class variation, resulting in degraded performance when confronted with previously unseen malware families. This paper presents MADOOD, a novel two stage, cluster driven deep learning framework for robust OOD malware detection and classification. In the first stage, malware family embeddings are modeled using class conditional spherical decision boundaries derived from Gaussian Discriminant Analysis (GDA), enabling statistically grounded separation of indistribution and OOD samples without requiring OOD data during training. Z score based distance analysis across multiple class centroids is employed to reliably identify anomalous samples in the latent space. In the second stage, a deep neural network integrates cluster based predictions, refined embeddings, and supervised classifier outputs to enhance final classification accuracy. Extensive evaluations on benchmark malware datasets comprising 25 known families and multiple novel OOD variants demonstrate that MADOOD significantly outperforms state of the art OOD detection methods, achieving an AUC of up to 0.911 on unseen malware families. The proposed framework provides a scalable, interpretable, and statistically principled solution for real world malware detection and anomaly identification in evolving cybersecurity environments.
【2】SafeBench-Seq: A Homology-Clustered, CPU-Only Baseline for Protein Hazard Screening with Physicochemical/Composition Features and Cluster-Aware Confidence Intervals
标题：SafeBench-Seq：具有物理化学/成分特征和检测器感知置信区间的蛋白质危害筛查的同质化、仅限PU基线
链接：https://arxiv.org/abs/2512.17527
摘要：蛋白质设计的基础模型提出了具体的生物安全风险，但社区缺乏一个简单的，可重复的基线序列水平的危险筛选，明确评估同源性控制和商品CPU上运行。我们介绍了SafeBench-Seq，这是一个仅元数据的，可重复的基准和基线分类器，完全基于公共数据（SafeProtein危害和UniProt benigns）和可解释的特征（全局理化描述符和氨基酸组成）。为了近似“前所未见”的威胁，我们以<=40%的同一性对组合数据集进行同源聚类，并执行聚类级保留（训练/测试之间没有聚类重叠）。我们报告了区分度（AUROC/AUPRC）和筛选操作点（TPR@1%FPR; FPR@95%TPR），具有95%自举置信区间（n=200），并通过CalibratedClassifierCV（逻辑回归/随机森林的等渗;线性SVM的Platt sigmoid）提供校准概率。我们量化概率质量使用Brier评分，预期校准误差（ECE; 15箱），和可靠性图。通过成分保留残留物重排和仅长度/成分消融来探测敏感性。从经验上讲，随机分裂大大高估了鲁棒性相对于同源聚类评价，校准的线性模型表现出比较好的校准，而树合奏保留略高的Brier/ECE。SafeBench-Seq是CPU专用的、可复制的，并且仅发布元数据（加入、集群ID、分割标签），从而实现严格的评估，而不会分发危险序列。
摘要：Foundation models for protein design raise concrete biosecurity risks, yet the community lacks a simple, reproducible baseline for sequence-level hazard screening that is explicitly evaluated under homology control and runs on commodity CPUs. We introduce SafeBench-Seq, a metadata-only, reproducible benchmark and baseline classifier built entirely from public data (SafeProtein hazards and UniProt benigns) and interpretable features (global physicochemical descriptors and amino-acid composition). To approximate "never-before-seen" threats, we homology-cluster the combined dataset at <=40% identity and perform cluster-level holdouts (no cluster overlap between train/test). We report discrimination (AUROC/AUPRC) and screening-operating points (TPR@1% FPR; FPR@95% TPR) with 95% bootstrap confidence intervals (n=200), and we provide calibrated probabilities via CalibratedClassifierCV (isotonic for Logistic Regression / Random Forest; Platt sigmoid for Linear SVM). We quantify probability quality using Brier score, Expected Calibration Error (ECE; 15 bins), and reliability diagrams. Shortcut susceptibility is probed via composition-preserving residue shuffles and length-/composition-only ablations. Empirically, random splits substantially overestimate robustness relative to homology-clustered evaluation; calibrated linear models exhibit comparatively good calibration, while tree ensembles retain slightly higher Brier/ECE. SafeBench-Seq is CPU-only, reproducible, and releases metadata only (accessions, cluster IDs, split labels), enabling rigorous evaluation without distributing hazardous sequences.
【3】Linear Attention for Joint Power Optimization and User-Centric Clustering in Cell-Free Networks
标题：无细胞网络中联合功率优化和以用户为中心的分簇的线性关注
链接：https://arxiv.org/abs/2512.17466
备注：Submitted
摘要：在以用户为中心的无小区大规模MIMO系统中，最优的AP分簇和功率分配是关键。现有的深度学习模型缺乏处理动态网络配置的灵活性。此外，许多方法忽略了导频污染，并且具有高计算复杂度。在本文中，我们提出了一个轻量级的Transformer模型，克服了这些限制，联合预测AP集群和功率仅从用户设备和AP的空间坐标。我们的模型是架构不可知的用户负载，处理集群和功率分配，而无需信道估计开销，并消除导频污染，通过分配用户到AP内的导频重用约束。我们还结合了一个定制的线性注意力机制，以有效地捕捉用户-AP交互，并实现线性可扩展性的用户数量。数值结果证实了该模型的有效性，在最大限度地提高最低频谱效率，并提供接近最佳的性能，同时确保在动态场景中的适应性和可扩展性。
摘要 ：Optimal AP clustering and power allocation are critical in user-centric cell-free massive MIMO systems. Existing deep learning models lack flexibility to handle dynamic network configurations. Furthermore, many approaches overlook pilot contamination and suffer from high computational complexity. In this paper, we propose a lightweight transformer model that overcomes these limitations by jointly predicting AP clusters and powers solely from spatial coordinates of user devices and AP. Our model is architecture-agnostic to users load, handles both clustering and power allocation without channel estimation overhead, and eliminates pilot contamination by assigning users to AP within a pilot reuse constraint. We also incorporate a customized linear attention mechanism to capture user-AP interactions efficiently and enable linear scalability with respect to the number of users. Numerical results confirm the model's effectiveness in maximizing the minimum spectral efficiency and providing near-optimal performance while ensuring adaptability and scalability in dynamic scenarios.
自动驾驶|车辆|车道检测等(2篇)
【1】Learning Safe Autonomous Driving Policies Using Predictive Safety Representations
标题：使用预测安全表示学习安全自动驾驶策略
链接：https://arxiv.org/abs/2512.17586
备注：8 pages, 4 figures. Submitted to ICRA 2026
摘要：安全强化学习（SafeRL）是自动驾驶的一个重要范例，要求智能体在严格的安全要求下优化性能。这种双重目标造成了根本性的紧张关系，因为过于保守的政策限制了驾驶效率，而激进的勘探则有违反安全规定的风险。安全策略学习的安全表示（SRPL）框架通过为代理提供未来约束违反的预测模型来解决这一挑战，并在受控环境中显示出希望。本文研究SRPL是否扩展到现实世界的自动驾驶场景。在Waymo Open Motion数据集（WOMD）和NuPlan上进行的系统实验表明，SRPL可以改善奖励-安全权衡，在成功率（效应大小r = 0.65-0.86）和成本降低（效应大小r = 0.70-0.83）方面实现统计学显著改善，观察到的改善p < 0.05。但是，其有效性取决于底层策略优化器和数据集分布。结果进一步表明，预测性安全表示在提高对观测噪声的鲁棒性方面起着关键作用。此外，在zero-shot交叉数据集评估中，SRPL增强代理与非SRPL方法相比表现出改进的泛化。这些发现共同证明了预测性安全表示在加强SafeRL自动驾驶方面的潜力。
摘要：Safe reinforcement learning (SafeRL) is a prominent paradigm for autonomous driving, where agents are required to optimize performance under strict safety requirements. This dual objective creates a fundamental tension, as overly conservative policies limit driving efficiency while aggressive exploration risks safety violations. The Safety Representations for Safer Policy Learning (SRPL) framework addresses this challenge by equipping agents with a predictive model of future constraint violations and has shown promise in controlled environments. This paper investigates whether SRPL extends to real-world autonomous driving scenarios. Systematic experiments on the Waymo Open Motion Dataset (WOMD) and NuPlan demonstrate that SRPL can improve the reward-safety tradeoff, achieving statistically significant improvements in success rate (effect sizes r = 0.65-0.86) and cost reduction (effect sizes r = 0.70-0.83), with p < 0.05 for observed improvements. However, its effectiveness depends on the underlying policy optimizer and the dataset distribution. The results further show that predictive safety representations play a critical role in improving robustness to observation noise. Additionally, in zero-shot cross-dataset evaluation, SRPL-augmented agents demonstrate improved generalization compared to non-SRPL methods. These findings collectively demonstrate the potential of predictive safety representations to strengthen SafeRL for autonomous driving.
【2】Electric Vehicle Charging Load Forecasting: An Experimental Comparison of Machine Learning Methods
标题：电动汽车充电负荷预测：机器学习方法的实验比较
链接：https://arxiv.org/abs/2512.17257
备注：18 pages, 2 figures, 5 tables
摘要：随着电动汽车作为应对气候变化的一种手段越来越受欢迎，人们开始关注其对电网管理的影响。因此，电动汽车充电需求预测成为一个及时而重要的研究课题。虽然大量的研究已经解决了交通能源负荷预测，相对较少的研究系统地比较多种预测方法在不同的时间范围和空间聚集水平在不同的城市环境。这项工作研究了五种时间序列预测模型的有效性，从传统的统计方法到机器学习和深度学习方法。预测性能评估短期，中期和长期范围（分别为分钟，小时和天的顺序），并跨越从单个充电站到区域和城市级聚合的空间尺度。该分析是在四个公开的真实世界数据集上进行的，每个数据集都独立报告了结果。据我们所知，这是第一项使用多个真实世界数据集在如此广泛的时间范围和空间聚合水平上系统评估电动汽车充电需求预测的工作。
摘要：With the growing popularity of electric vehicles as a means of addressing climate change, concerns have emerged regarding their impact on electric grid management. As a result, predicting EV charging demand has become a timely and important research problem. While substantial research has addressed energy load forecasting in transportation, relatively few studies systematically compare multiple forecasting methods across different temporal horizons and spatial aggregation levels in diverse urban settings. This work investigates the effectiveness of five time series forecasting models, ranging from traditional statistical approaches to machine learning and deep learning methods. Forecasting performance is evaluated for short-, mid-, and long-term horizons (on the order of minutes, hours, and days, respectively), and across spatial scales ranging from individual charging stations to regional and city-level aggregations. The analysis is conducted on four publicly available real-world datasets, with results reported independently for each dataset. To the best of our knowledge, this is the first work to systematically evaluate EV charging demand forecasting across such a wide range of temporal horizons and spatial aggregation levels using multiple real-world datasets.
联邦学习|隐私保护|加密(1篇)
【1】Practical Framework for Privacy-Preserving and Byzantine-robust Federated Learning
标题：隐私保护和拜占庭稳健联邦学习的实用框架
链接：https://arxiv.org/abs/2512.17254
备注：Accepted for publication in IEEE Transactions on Information Forensics and Security
摘要：联合学习（FL）允许多个客户端协作训练模型，而无需共享其私有数据。然而，FL容易受到拜占庭攻击，其中对手操纵客户端模型以损害联邦模型，以及隐私推断攻击，其中对手利用客户端模型来推断私有数据。针对后门和隐私推断攻击的现有防御引入了显著的计算和通信开销，在理论和实践之间产生了差距。为了解决这个问题，我们提出了ABBR，一个实用的框架拜占庭鲁棒和隐私保护FL。我们是第一个利用降维来加速隐私保护FL中复杂过滤规则的私有计算。此外，我们分析了在低-维空间，并引入自适应调整策略，以最大限度地减少绕过过滤的恶意模型对全局模型的影响。我们使用最先进的拜占庭鲁棒聚合规则实现了ABBR，并在公共数据集上对其进行了评估，结果表明它运行速度更快，通信开销最小，并且保持了与基线几乎相同的拜占庭弹性。
摘要：Federated Learning (FL) allows multiple clients to collaboratively train a model without sharing their private data. However, FL is vulnerable to Byzantine attacks, where adversaries manipulate client models to compromise the federated model, and privacy inference attacks, where adversaries exploit client models to infer private data. Existing defenses against both backdoor and privacy inference attacks introduce significant computational and communication overhead, creating a gap between theory and practice. To address this, we propose ABBR, a practical framework for Byzantine-robust and privacy-preserving FL. We are the first to utilize dimensionality reduction to speed up the private computation of complex filtering rules in privacy-preserving FL. Additionally, we analyze the accuracy loss of vector-wise filtering in low-dimensional space and introduce an adaptive tuning strategy to minimize the impact of malicious models that bypass filtering on the global model. We implement ABBR with state-of-the-art Byzantine-robust aggregation rules and evaluate it on public datasets, showing that it runs significantly faster, has minimal communication overhead, and maintains nearly the same Byzantine-resilience as the baselines.
推理|分析|理解|解释(8篇)
【1】Machine Learning for Static and Single-Event Dynamic Complex Network Analysis
标题：静态和单事件动态复杂网络分析的机器学习
链接：https://arxiv.org/abs/2512.17577
摘要：本论文的主要目标是为静态和单事件动态网络的图表示学习开发新的算法方法。在这个方向上，我们专注于潜在空间模型家族，更具体地说，是潜在距离模型，它自然地传达了重要的网络特征，如同质性，传递性和平衡理论。此外，本论文的目的是创建结构感知的网络表示，导致网络结构的层次化表达，社区特征，网络中的极端配置文件的识别，以及时间网络中的影响动力学量化。至关重要的是，所提出的方法旨在定义统一的学习过程，消除了对后处理步骤等多阶段过程的需求。我们的目标是深入研究统一网络嵌入的旅程，这些网络嵌入既全面又强大，能够表征网络结构，并熟练地处理图分析提供的各种任务。
摘要：The primary objective of this thesis is to develop novel algorithmic approaches for Graph Representation Learning of static and single-event dynamic networks. In such a direction, we focus on the family of Latent Space Models, and more specifically on the Latent Distance Model which naturally conveys important network characteristics such as homophily, transitivity, and the balance theory. Furthermore, this thesis aims to create structural-aware network representations, which lead to hierarchical expressions of network structure, community characterization, the identification of extreme profiles in networks, and impact dynamics quantification in temporal networks. Crucially, the methods presented are designed to define unified learning processes, eliminating the need for heuristics and multi-stage processes like post-processing steps. Our aim is to delve into a journey towards unified network embeddings that are both comprehensive and powerful, capable of characterizing network structures and adeptly handling the diverse tasks that graph analysis offers.
【2】Enabling Disaggregated Multi-Stage MLLM Inference via GPU-Internal Scheduling and Resource Sharing
标题：通过GPU内部调度和资源共享启用分散的多阶段MLLM推理
链接：https://arxiv.org/abs/2512.17574
摘要：多模态大型语言模型（MLLM）通过三个阶段的管道扩展了LLM的视觉理解：多模态预处理，视觉编码和LLM推理。虽然这些阶段增强了能力，但它们引入了重大的系统瓶颈。首先，多模式预处理，特别是视频解码，通常占主导地位的时间到第一个令牌（TTFT）。大多数系统依赖于基于CPU的解码，这严重限制了吞吐量，而现有的基于GPU的方法优先考虑面向吞吐量的并行性，无法满足MLLM推理的延迟敏感要求。其次，视觉编码器是一个独立的、计算密集型的阶段，它产生视觉嵌入，不能与LLM预填充或解码协同处理。这种异构性会导致阶段间阻塞，并增加令牌生成延迟。即使部署在单独的GPU上，这些阶段也无法充分利用可用的计算和内存资源，从而降低了整体利用率并限制了系统吞吐量。 为了应对这些挑战，我们提出了FlashCodec和UnifiedServe，这两种互补的设计可以共同优化端到端MLLM管道。FlashCodec通过协作式多GPU视频解码加速多模式预处理阶段，在保持高吞吐量的同时减少解码延迟。UnifiedServe使用逻辑解耦的执行来优化视觉到文本和推理阶段，以消除阶段间的阻塞，但物理上共享GPU资源，以最大限度地提高GPU系统利用率。通过仔细编排跨阶段的执行并最大限度地减少干扰，UnifiedServe Together，我们提出的框架形成了一个端到端的优化堆栈，可以服务多达3.0$\times$更多的请求或强制执行1.5$\times$更严格的SLO，同时实现高达4.4$\times$更高的吞吐量与最先进的系统相比。
摘要：Multimodal large language models (MLLMs) extend LLMs with visual understanding through a three-stage pipeline: multimodal preprocessing, vision encoding, and LLM inference. While these stages enhance capability, they introduce significant system bottlenecks. First, multimodal preprocessing-especially video decoding-often dominates Time-to-First-Token (TTFT). Most systems rely on CPU-based decoding, which severely limits throughput, while existing GPU-based approaches prioritize throughput-oriented parallelism and fail to meet the latency-sensitive requirements of MLLM inference. Second, the vision encoder is a standalone, compute-intensive stage that produces visual embeddings and cannot be co-batched with LLM prefill or decoding. This heterogeneity forces inter-stage blocking and increases token-generation latency. Even when deployed on separate GPUs, these stages underutilize available compute and memory resources, reducing overall utilization and constraining system throughput. To address these challenges, we present FlashCodec and UnifiedServe, two complementary designs that jointly optimize the end-to-end MLLM pipeline. FlashCodec accelerates the multimodal preprocessing stage through collaborative multi-GPU video decoding, reducing decoding latency while preserving high throughput. UnifiedServe optimizes the vision-to-text and inference stages using a logically decoupled their execution to eliminate inter-stage blocking, yet physically sharing GPU resources to maximize GPU system utilization. By carefully orchestrating execution across stages and minimizing interference, UnifiedServe Together, our proposed framework forms an end-to-end optimized stack that can serve up to 3.0$\times$ more requests or enforce 1.5$\times$ tighter SLOs, while achieving up to 4.4$\times$ higher throughput compared to state-of-the-art systems.
【3】Learning What to Write: Write-Gated KV for Efficient Long-Context Inference
标题：学习写什么：写门控GV以实现高效的长上下文推理
链接：https://arxiv.org/abs/2512.17452
摘要：长上下文LLM推理受到二次注意复杂度和线性KV缓存增长的检验。先前的方法通过事后选择或驱逐来缓解这一点，但忽略了根本的低效率：不加选择地写入持久存储器。在本文中，我们将KV缓存管理形式化为三个基元的因果系统：KV接纳、选择和驱逐。我们通过写门KV实例化KV准入，这是一种轻量级机制，可以在令牌进入缓存之前学习预测令牌实用程序。通过尽早过滤掉低效用状态以保持紧凑的全局缓存以及滑动的本地缓存，写门控KV将内存使用量减少了46-57%，并在Llama模型上提供了3.03-3.45$\times$预填充和1.89-2.56$\times$解码加速，精度损失可以忽略不计，同时保持与FlashAttention和分页KV系统的兼容性。这些结果表明，学习写什么是有效的长上下文推理的原则和实用的配方。代码可在https://github.com/EMCLab-Sinica/WG-KV上获得。
摘要：Long-context LLM inference is bottlenecked by the quadratic attention complexity and linear KV cache growth. Prior approaches mitigate this via post-hoc selection or eviction but overlook the root inefficiency: indiscriminate writing to persistent memory. In this paper, we formalize KV cache management as a causal system of three primitives: KV Admission, Selection, and Eviction. We instantiate KV Admission via Write-Gated KV, a lightweight mechanism that learns to predict token utility before it enters the cache. By filtering out low-utility states early to maintain a compact global cache alongside a sliding local cache, Write-Gated KV reduces memory usage by 46-57% and delivers 3.03-3.45$\times$ prefill and 1.89-2.56$\times$ decode speedups on Llama model with negligible accuracy loss, all while remaining compatible with FlashAttention and paged-KV systems. These results demonstrate that learning what to write, is a principled and practical recipe for efficient long-context inference. Code is available at https://github.com/EMCLab-Sinica/WG-KV .
【4】meval: A Statistical Toolbox for Fine-Grained Model Performance Analysis
标题：meval：细粒度模型性能分析的统计表
链接：https://arxiv.org/abs/2512.17409
摘要：分析按患者和记录属性分层的机器学习模型性能正在成为公认的规范，并且通常会产生有关重要模型故障模式的重要见解。然而，以统计学上严格的方式进行这样的分析是不平凡的。必须选择适当的业绩衡量标准，以便在不同样本量和基本比率的群体之间进行有效的比较;必须确定衡量标准的不确定性，并对多重比较进行校正，以评估观察到的任何差异是否纯粹是偶然造成的;在交叉分析的情况下，必须实施机制以在组合的许多子组组合中找到最“感兴趣的”子组。我们在这里提出了一个统计工具箱，解决这些挑战，使从业者能够轻松而严格地评估他们的模型，潜在的子组性能差异。虽然广泛适用，但该工具箱专为医学成像应用而设计。工具箱提供的分析在两个案例研究中得到了说明，一个是ISIC 2020数据集上的皮肤病变恶性肿瘤分类，另一个是MIMIC-CXR数据集上基于胸部X射线的疾病分类。
摘要 ：Analyzing machine learning model performance stratified by patient and recording properties is becoming the accepted norm and often yields crucial insights about important model failure modes. Performing such analyses in a statistically rigorous manner is non-trivial, however. Appropriate performance metrics must be selected that allow for valid comparisons between groups of different sample sizes and base rates; metric uncertainty must be determined and multiple comparisons be corrected for, in order to assess whether any observed differences may be purely due to chance; and in the case of intersectional analyses, mechanisms must be implemented to find the most `interesting' subgroups within combinatorially many subgroup combinations. We here present a statistical toolbox that addresses these challenges and enables practitioners to easily yet rigorously assess their models for potential subgroup performance disparities. While broadly applicable, the toolbox is specifically designed for medical imaging applications. The analyses provided by the toolbox are illustrated in two case studies, one in skin lesion malignancy classification on the ISIC2020 dataset and one in chest X-ray-based disease classification on the MIMIC-CXR dataset.
【5】DeepShare: Sharing ReLU Across Channels and Layers for Efficient Private Inference
标题：DeepShare：跨渠道和层共享ReLU，以实现高效的私人推理
链接：https://arxiv.org/abs/2512.17398
摘要：私有推理（PI）使用加密原语来执行隐私保护机器学习。在这种情况下，网络的所有者对客户端的数据进行推理，而不了解任何有关数据的信息，也不透露有关模型的任何信息。已经观察到PI的主要计算瓶颈是门的计算（即，ReLU），因此已经投入了大量的精力来减少给定网络中的ReLU数量。 我们专注于DReLU，它是ReLU的非线性阶跃函数，并表明一个DReLU可以服务于许多ReLU操作。我们提出了一个新的激活模块，其中DReLU操作仅在通道的子集（原型通道）上执行，而其余通道（复制通道）从原型通道中的对应神经元复制其每个神经元的DReLU。然后，我们将这个想法扩展到不同的层。 我们表明，这种公式可以大大减少resnet类型网络中DReLU操作的数量。此外，我们的理论分析表明，这种新的配方可以解决扩展版本的XOR问题，只使用一个非线性和两个神经元，这是传统的配方和一些PI特定的方法无法实现的。我们在几个分类设置上实现了新的SOTA结果，并在图像分割上实现了SOTA结果。
摘要：Private Inference (PI) uses cryptographic primitives to perform privacy preserving machine learning. In this setting, the owner of the network runs inference on the data of the client without learning anything about the data and without revealing any information about the model. It has been observed that a major computational bottleneck of PI is the calculation of the gate (i.e., ReLU), so a considerable amount of effort have been devoted to reducing the number of ReLUs in a given network. We focus on the DReLU, which is the non-linear step function of the ReLU and show that one DReLU can serve many ReLU operations. We suggest a new activation module where the DReLU operation is only performed on a subset of the channels (Prototype channels), while the rest of the channels (replicate channels) replicates the DReLU of each of their neurons from the corresponding neurons in one of the prototype channels. We then extend this idea to work across different layers. We show that this formulation can drastically reduce the number of DReLU operations in resnet type network. Furthermore, our theoretical analysis shows that this new formulation can solve an extended version of the XOR problem, using just one non-linearity and two neurons, something that traditional formulations and some PI specific methods cannot achieve. We achieve new SOTA results on several classification setups, and achieve SOTA results on image segmentation.
【6】Understanding Generalization in Role-Playing Models via Information Theory
标题：通过信息理论理解角色扮演模型中的概括
链接：https://arxiv.org/abs/2512.17270
摘要：角色扮演模型（RPMs）广泛用于现实世界的应用程序中，但在野外部署时表现不佳。这种退化可以归因于分布变化，包括用户、角色和对话成分的变化。现有的方法，如LLM-as-a-judge在提供这些变化如何影响RPM泛化的细粒度诊断方面存在不足，因此缺乏正式的框架来表征RPM泛化行为。为了弥合这些差距，我们引入了一个信息理论的度量，命名为基于推理的有效互信息差（R-EMID），以衡量RPM性能下降的可解释的方式。我们还推导出一个上界的R-EMID预测RPM的最坏情况下的泛化性能，并从理论上揭示了各种变化如何有助于RPM的性能下降。此外，我们提出了一个协同进化的强化学习框架，以自适应地建模用户，角色和对话上下文之间的连接，从而提高对话响应生成概率的估计，这是计算R-EMID的关键。最后，我们使用R-EMID评估了各种RPM的泛化性能，发现用户转变在所有转变中构成的风险最高，强化学习是增强RPM泛化的最有效方法。
摘要：Role-playing models (RPMs) are widely used in real-world applications but underperform when deployed in the wild. This degradation can be attributed to distribution shifts, including user, character, and dialogue compositional shifts. Existing methods like LLM-as-a-judge fall short in providing a fine-grained diagnosis of how these shifts affect RPM generalization, and thus there lack formal frameworks to characterize RPM generalization behaviors. To bridge these gaps, we introduce an information-theoretic metric, named reasoning-based effective mutual information difference (R-EMID), to measure RPM performance degradation in an interpretable way. We also derive an upper bound on R-EMID to predict the worst-case generalization performance of RPMs and theoretically reveal how various shifts contribute to the RPM performance degradation. Moreover, we propose a co-evolving reinforcement learning framework to adaptively model the connection among user, character, and dialogue context and thus enhance the estimation of dialogue response generation probability, which is critical for calculating R-EMID. Finally, we evaluate the generalization performance of various RPMs using R-EMID, finding that user shift poses the highest risk among all shifts and reinforcement learning is the most effective approach for enhancing RPM generalization.
【7】A Theoretical Analysis of State Similarity Between Markov Decision Processes
标题：马尔科夫决策过程状态相似性的理论分析
链接：https://arxiv.org/abs/2512.17265
备注：Submitted to an IEEE Transactions. arXiv admin note: substantial text overlap with arXiv:2509.18714
摘要：互模拟度量（BSM）是分析马尔可夫决策过程（MDP）中状态相似性的有力工具，它揭示了在BSM中越接近的状态具有越相似的最优值函数。虽然BSM已经成功地用于强化学习（RL）中的状态表示学习和策略探索等任务，但它在多个MDP之间的状态相似性方面的应用仍然具有挑战性。先前的工作试图将BSM扩展到MDP对，但是缺乏完善的数学性质限制了MDP之间的进一步理论分析。在这项工作中，我们正式建立了一个广义互模拟度量（GBSM）用于测量任意MDP对之间的状态相似性，这是严格证明了三个基本度量属性，即，GBSM对称性，内MDP三角不等式，以及恒等空间上的距离界。利用这些属性，我们从理论上分析的政策转移，状态聚合，并在MDP基于采样的估计，获得明确的界限，严格比现有的标准BSM。此外，GBSM提供了一个封闭形式的样本复杂性估计，改善现有的渐近结果的基础上BSM。数值结果验证了我们的理论研究结果，并证明了GBSM在多MDP场景的有效性。
摘要 ：The bisimulation metric (BSM) is a powerful tool for analyzing state similarities within a Markov decision process (MDP), revealing that states closer in BSM have more similar optimal value functions. While BSM has been successfully utilized in reinforcement learning (RL) for tasks like state representation learning and policy exploration, its application to state similarity between multiple MDPs remains challenging. Prior work has attempted to extend BSM to pairs of MDPs, but a lack of well-established mathematical properties has limited further theoretical analysis between MDPs. In this work, we formally establish a generalized bisimulation metric (GBSM) for measuring state similarity between arbitrary pairs of MDPs, which is rigorously proven with three fundamental metric properties, i.e., GBSM symmetry, inter-MDP triangle inequality, and a distance bound on identical spaces. Leveraging these properties, we theoretically analyze policy transfer, state aggregation, and sampling-based estimation across MDPs, obtaining explicit bounds that are strictly tighter than existing ones derived from the standard BSM. Additionally, GBSM provides a closed-form sample complexity for estimation, improving upon existing asymptotic results based on BSM. Numerical results validate our theoretical findings and demonstrate the effectiveness of GBSM in multi-MDP scenarios.
【8】Can Large Reasoning Models Improve Accuracy on Mathematical Tasks Using Flawed Thinking?
标题：大型推理模型能否使用有缺陷的思维提高数学任务的准确性？
链接：https://arxiv.org/abs/2512.17079
摘要：思想链（CoT）提示已经成为大型语言模型中数学推理的核心，但模型仍然容易出现早期错误：单个算术失误或不合理的推理通常会传播到不正确的最终答案。我们研究了在故意有缺陷的推理痕迹上进行训练是否可以教会模型在不降低标准问题解决能力的情况下检测并从这些错误中恢复。使用MATH-lighteval中的竞赛级问题，我们生成包含一个控制错误的CoT前缀，无论是计算错误（符号翻转，丢失的术语）还是推理错误（误用的规则，不合理的逻辑步骤），并使用二进制最终答案奖励使用GRPO微调Qwen 3 - 4 B。我们的Mixed-CoT-RL模型在干净的问题上与标准RL相匹配（41%对41%），而在预填充有缺陷推理的问题上则表现出色（24%对19%）。值得注意的是，仅清洁的RL微调将鲁棒性降低到未调整的基线以下（19%对20%），这表明传统训练增加了对误导性预填充的敏感性。在错误类型中，对推理错误的训练比单独的计算错误产生更大的鲁棒性增益，混合训练表现最好。这些发现表明，在训练过程中暴露于有缺陷的痕迹可以在不牺牲准确性的情况下改善错误恢复行为，这表明LLM中有一条更强大的数学推理之路。
摘要：Chain-of-thought (CoT) prompting has become central to mathematical reasoning in large language models, yet models remain brittle to early errors: a single arithmetic slip or unjustified inference typically propagates uncorrected to an incorrect final answer. We investigate whether training on intentionally flawed reasoning traces can teach models to detect and recover from such errors without degrading standard problem-solving ability. Using competition-level problems from MATH-lighteval, we generate CoT prefixes containing exactly one controlled error, either a calculation error (sign flips, dropped terms) or a reasoning error (misapplied rules, unjustified logical steps), and fine-tune Qwen3-4B with GRPO using a binary final-answer reward. Our Mixed-CoT-RL model matches standard RL on clean problems (41% vs 41%) while substantially outperforming it on problems prefilled with flawed reasoning (24% vs 19%). Notably, clean-only RL fine-tuning degrades robustness below the untuned baseline 19% vs. 20%), indicating that conventional training increases susceptibility to misleading prefills. Among error types, training on reasoning errors yields greater robustness gains than calculation errors alone, with mixed training performing best. These findings demonstrate that exposure to flawed traces during training can improve error-recovery behavior without sacrificing accuracy, suggesting a path toward more robust mathematical reasoning in LLMs.
检测相关(3篇)
【1】Adversarially Robust Detection of Harmful Online Content: A Computational Design Science Approach
标题：有害在线内容的对抗稳健检测：计算设计科学方法
链接：https://arxiv.org/abs/2512.17367
摘要：社交媒体平台受到仇恨言论、错误信息和极端主义言论等有害内容的困扰。机器学习（ML）模型被广泛用于检测此类内容;然而，它们仍然非常容易受到对抗性攻击，其中恶意用户巧妙地修改文本以逃避检测。因此，增强对抗性鲁棒性至关重要，需要检测器能够抵御各种攻击（泛化能力），同时保持高的整体准确性。然而，同时实现最佳的概括性和准确性是具有挑战性的。遵循计算设计科学范式，本研究采用顺序方法，首先通过识别文本对抗性攻击的关键不变性并利用它们来确保检测器实例化，提出了一种新型框架（基于大型语言模型的样本生成和聚合，LLM-SGA）框架内具有很强的概括性。其次，我们实例化我们的检测器（Adversarially Robust Harmful Online Content Detector，ARHOCD），其具有三个新颖的设计组件以提高检测精度：（1）利用它们的互补强度的多个碱基检测器的集合;（2）基于每个样本的可预测性和每个碱基检测器的能力动态调整权重的新颖权重分配方法，权重使用领域知识初始化并通过贝叶斯推理更新;以及（3）一种新的对抗训练策略，迭代优化基本检测器和权重分配器。我们解决了现有对抗性鲁棒性增强研究的几个局限性，并在仇恨言论、谣言和极端主义内容的三个数据集上对ARHOCD进行了实证评估。实验结果表明，ARHOCD算法具有较强的泛化能力，提高了对抗条件下的检测精度。
摘要：Social media platforms are plagued by harmful content such as hate speech, misinformation, and extremist rhetoric. Machine learning (ML) models are widely adopted to detect such content; however, they remain highly vulnerable to adversarial attacks, wherein malicious users subtly modify text to evade detection. Enhancing adversarial robustness is therefore essential, requiring detectors that can defend against diverse attacks (generalizability) while maintaining high overall accuracy. However, simultaneously achieving both optimal generalizability and accuracy is challenging. Following the computational design science paradigm, this study takes a sequential approach that first proposes a novel framework (Large Language Model-based Sample Generation and Aggregation, LLM-SGA) by identifying the key invariances of textual adversarial attacks and leveraging them to ensure that a detector instantiated within the framework has strong generalizability. Second, we instantiate our detector (Adversarially Robust Harmful Online Content Detector, ARHOCD) with three novel design components to improve detection accuracy: (1) an ensemble of multiple base detectors that exploits their complementary strengths; (2) a novel weight assignment method that dynamically adjusts weights based on each sample's predictability and each base detector's capability, with weights initialized using domain knowledge and updated via Bayesian inference; and (3) a novel adversarial training strategy that iteratively optimizes both the base detectors and the weight assignor. We addressed several limitations of existing adversarial robustness enhancement research and empirically evaluated ARHOCD across three datasets spanning hate speech, rumor, and extremist content. Results show that ARHOCD offers strong generalizability and improves detection accuracy under adversarial conditions.
【2】LibriVAD: A Scalable Open Dataset with Deep Learning Benchmarks for Voice Activity Detection
标题：LibriVAR：具有深度学习基准的可扩展开放数据集，用于语音活动检测
链接：https://arxiv.org/abs/2512.17281
摘要：鲁棒的语音活动检测（VAD）仍然是一项具有挑战性的任务，特别是在嘈杂，多样化和不可见的声学条件下。除了算法开发之外，推进VAD研究的一个关键限制是缺乏大规模，系统控制和公开可用的数据集。为了解决这个问题，我们引入了LibriVAD -一个可扩展的开源数据集，来自LibriSpeech，并增加了各种真实世界和合成噪声源。LibriVAD能够系统地控制语音噪声比、静音语音比（SSR）和噪声多样性，并以三种大小（15 GB、150 GB和1.5 TB）发布，其中两种变体（LibriVAD-NonConcat和LibriVAD-Concat）支持不同的实验设置。我们对多个特征模型组合进行基准测试，包括波形、Mel频率倒谱系数（MFCC）和Gammatone滤波器组倒谱系数，并介绍了用于VAD的Vision Transformer（ViT）架构。我们的实验表明，具有MFCC特征的ViT在可见、不可见和分布外（OOD）条件下的性能始终优于已建立的VAD模型，如增强型深度神经网络和卷积长短期记忆深度神经网络，包括对真实世界VOiCES数据集的评估。我们进一步分析了数据集大小和SSR对模型泛化的影响，实验表明，在OOD条件下，扩大数据集大小和平衡SSR显著且一致地增强了VAD性能。所有数据集、训练模型和代码都公开发布，以促进可重复性并加速VAD研究的进展。
摘要 ：Robust Voice Activity Detection (VAD) remains a challenging task, especially under noisy, diverse, and unseen acoustic conditions. Beyond algorithmic development, a key limitation in advancing VAD research is the lack of large-scale, systematically controlled, and publicly available datasets. To address this, we introduce LibriVAD - a scalable open-source dataset derived from LibriSpeech and augmented with diverse real-world and synthetic noise sources. LibriVAD enables systematic control over speech-to-noise ratio, silence-to-speech ratio (SSR), and noise diversity, and is released in three sizes (15 GB, 150 GB, and 1.5 TB) with two variants (LibriVAD-NonConcat and LibriVAD-Concat) to support different experimental setups. We benchmark multiple feature-model combinations, including waveform, Mel-Frequency Cepstral Coefficients (MFCC), and Gammatone filter bank cepstral coefficients, and introduce the Vision Transformer (ViT) architecture for VAD. Our experiments show that ViT with MFCC features consistently outperforms established VAD models such as boosted deep neural network and convolutional long short-term memory deep neural network across seen, unseen, and out-of-distribution (OOD) conditions, including evaluation on the real-world VOiCES dataset. We further analyze the impact of dataset size and SSR on model generalization, experimentally showing that scaling up dataset size and balancing SSR noticeably and consistently enhance VAD performance under OOD conditions. All datasets, trained models, and code are publicly released to foster reproducibility and accelerate progress in VAD research.
【3】Fraud detection in credit card transactions using Quantum-Assisted Restricted Boltzmann Machines
标题：使用量子辅助限制Boltzmann机进行信用卡交易中的欺诈检测
链接：https://arxiv.org/abs/2512.17660
备注：8 pages, 3 figures
摘要：随着量子计算机的处理效率和可用性的提高，新兴量子计算平台的用例变得具有经济意义。我们评估了量子计算辅助下的受限玻尔兹曼机（RBM）的性能，在真实的量子硬件和模拟器上运行，使用包含由巴西领先的金融科技公司Stone提供的1.45亿笔交易的真实数据集进行信用卡欺诈检测。结果表明，量子辅助RBM方法是能够实现优越的性能，在大多数的品质因数相比，经典的方法，即使使用当前的噪声量子退火。我们的研究为实现量子辅助RBM在金融系统中的一般故障检测铺平了道路。
摘要：Use cases for emerging quantum computing platforms become economically relevant as the efficiency of processing and availability of quantum computers increase. We assess the performance of Restricted Boltzmann Machines (RBM) assisted by quantum computing, running on real quantum hardware and simulators, using a real dataset containing 145 million transactions provided by Stone, a leading Brazilian fintech, for credit card fraud detection. The results suggest that the quantum-assisted RBM method is able to achieve superior performance in most figures of merit in comparison to classical approaches, even using current noisy quantum annealers. Our study paves the way for implementing quantum-assisted RBMs for general fault detection in financial systems.
分类|识别(1篇)
【1】QSMOTE-PGM/kPGM: QSMOTE Based PGM and kPGM for Imbalanced Dataset Classification
标题：QSMOTE-CGM/kCGM：基于QSMOTE的CGM和kCGM用于不平衡数据集分类
链接：https://arxiv.org/abs/2512.16960
备注：14 pages, 10 figures
摘要：量子启发机器学习（QiML）利用量子理论的数学框架来增强经典算法，特别强调高维特征空间中的内积结构。其中突出的方法，核技巧，广泛用于支持向量机，提供了有效的相似性计算，而相当好的测量（PGM），源于量子态歧视，使分类接地希尔伯特空间几何。基于核PGM（KPGM）和直接基于PGM的分类器的最新发展，这项工作提出了一个统一的理论和经验比较这些范例。我们使用量子SMOTE（QSMOTE）变体分析了它们在合成过采样场景中的性能。实验结果表明，PGM和KPGM分类器始终优于经典的随机森林基线，特别是当使用多个量子副本时。值得注意的是，具有立体声编码和n_copies=2的PGM实现了最高的总体准确度（0.8512）和F1得分（0.8234），而KPGM在QSMOTE变体中表现出竞争性和更稳定的行为，最高得分为0.8511（立体声）和0.8483（振幅）。这些发现强调了量子启发分类器不仅在召回和平衡性能方面提供了切实的收益，而且还提供了互补的优势：PGM受益于编码特定的增强，而KPGM确保了整个采样策略的鲁棒性。我们的研究结果促进了对基于内核和基于测量的QiML方法的理解，为它们在不同数据特征和计算约束下的适用性提供了实用指导。
摘要：Quantum-inspired machine learning (QiML) leverages mathematical frameworks from quantum theory to enhance classical algorithms, with particular emphasis on inner product structures in high-dimensional feature spaces. Among the prominent approaches, the Kernel Trick, widely used in support vector machines, provides efficient similarity computation, while the Pretty Good Measurement (PGM), originating from quantum state discrimination, enables classification grounded in Hilbert space geometry. Building on recent developments in kernelized PGM (KPGM) and direct PGM-based classifiers, this work presents a unified theoretical and empirical comparison of these paradigms. We analyze their performance across synthetic oversampling scenarios using Quantum SMOTE (QSMOTE) variants. Experimental results show that both PGM and KPGM classifiers consistently outperform a classical random forest baseline, particularly when multiple quantum copies are employed. Notably, PGM with stereo encoding and n_copies=2 achieves the highest overall accuracy (0.8512) and F1-score (0.8234), while KPGM demonstrates competitive and more stable behavior across QSMOTE variants, with top scores of 0.8511 (stereo) and 0.8483 (amplitude). These findings highlight that quantum-inspired classifiers not only provide tangible gains in recall and balanced performance but also offer complementary strengths: PGM benefits from encoding-specific enhancements, whereas KPGM ensures robustness across sampling strategies. Our results advance the understanding of kernel-based and measurement-based QiML methods, offering practical guidance on their applicability under varying data characteristics and computational constraints.
表征(2篇)
【1】A Unified Representation of Neural Networks Architectures
标题：神经网络架构的统一表示
链接：https://arxiv.org/abs/2512.17593
摘要：在本文中，我们考虑神经网络（NN）架构的极限情况下，在每个隐藏层的神经元的数量和隐藏层的数量趋于无穷大，从而形成一个连续的，我们推导出近似误差作为神经元和/或隐藏层的数量的函数。首先，我们考虑具有单个隐藏层的神经网络的情况，并推导出一个积分无限宽度神经表示，该表示概括了现有的连续神经网络（CNN）表示。然后，我们将其扩展到具有有限数量的整数隐藏层和剩余连接的深度剩余CNN。其次，我们重新审视神经常微分方程和深层残差神经网络之间的关系，并通过离散化技术形式化逼近误差。然后，我们将这两种方法合并为一个统一的均匀表示的神经网络作为一个分布参数神经网络（DiPaNet），我们表明，大多数现有的有限和无限维的神经网络架构是相关的通过离散化/离散化与DiPaNet表示。我们的方法是纯粹的确定性和适用于一般的，一致连续的矩阵权重函数。与神经领域的差异和相似之处进行了讨论，以及进一步可能的推广和应用的DiPaNet框架。
摘要：In this paper we consider the limiting case of neural networks (NNs) architectures when the number of neurons in each hidden layer and the number of hidden layers tend to infinity thus forming a continuum, and we derive approximation errors as a function of the number of neurons and/or hidden layers. Firstly, we consider the case of neural networks with a single hidden layer and we derive an integral infinite width neural representation that generalizes existing continuous neural networks (CNNs) representations. Then we extend this to deep residual CNNs that have a finite number of integral hidden layers and residual connections. Secondly, we revisit the relation between neural ODEs and deep residual NNs and we formalize approximation errors via discretization techniques. Then, we merge these two approaches into a unified homogeneous representation of NNs as a Distributed Parameter neural Network (DiPaNet) and we show that most of the existing finite and infinite-dimensional NNs architectures are related via homogeneization/discretization with the DiPaNet representation. Our approach is purely deterministic and applies to general, uniformly continuous matrix weight functions. Differences and similarities with neural fields are discussed along with further possible generalizations and applications of the DiPaNet framework.
【2】Disentangled representations via score-based variational autoencoders
标题：通过基于分数的变分自动编码器解纠缠表示
链接：https://arxiv.org/abs/2512.17127
备注：34 pages, 7 figures
摘要：我们提出了基于分数的多尺度推理自动编码器（SAMI），这是一种结合扩散模型和VAE理论框架的无监督表示学习方法。通过统一各自的证据下限，SAMI制定了一个原则性的目标，通过基于分数的指导下的基本扩散过程中学习表示。由此产生的表示自动捕获数据中有意义的结构：它在我们的合成数据集中恢复地面真值生成因子，从复杂的自然图像中学习因子分解的语义潜在维度，并将视频序列编码为比其他编码器更直的潜在轨迹，尽管只在静态图像上进行训练。此外，SAMI可以从预先训练的扩散模型中提取有用的表示，只需最少的额外训练。最后，显式概率公式提供了新的方法来识别语义上有意义的轴在没有监督的标签，其数学精确性使我们能够对学习表示的性质作出正式的声明。总体而言，这些结果表明，隐式的结构信息扩散模型可以明确和解释，通过协同组合与变分自动编码器。
摘要：We present the Score-based Autoencoder for Multiscale Inference (SAMI), a method for unsupervised representation learning that combines the theoretical frameworks of diffusion models and VAEs. By unifying their respective evidence lower bounds, SAMI formulates a principled objective that learns representations through score-based guidance of the underlying diffusion process. The resulting representations automatically capture meaningful structure in the data: it recovers ground truth generative factors in our synthetic dataset, learns factorized, semantic latent dimensions from complex natural images, and encodes video sequences into latent trajectories that are straighter than those of alternative encoders, despite training exclusively on static images. Furthermore, SAMI can extract useful representations from pre-trained diffusion models with minimal additional training. Finally, the explicitly probabilistic formulation provides new ways to identify semantically meaningful axes in the absence of supervised labels, and its mathematical exactness allows us to make formal statements about the nature of the learned representation. Overall, these results indicate that implicit structural information in diffusion models can be made explicit and interpretable through synergistic combination with a variational autoencoder.
3D|3D重建等相关(1篇)
【1】DiffeoMorph: Learning to Morph 3D Shapes Using Differentiable Agent-Based Simulations
标题：DeliveroMorph：学习使用基于可区分代理的模拟变形3D收件箱
链接：https://arxiv.org/abs/2512.17129
摘要：生物系统可以通过相同代理的集体行为形成复杂的三维结构-遵循相同内部规则并在没有中央控制的情况下进行通信的细胞。这种分布式控制如何产生精确的全球模式仍然是一个中心问题，不仅在发育生物学，而且在分布式机器人，可编程物质和多智能体学习。在这里，我们介绍了一个端到端的可区分框架，用于学习一个形态生成协议，该协议指导一群代理变形为目标3D形状。每个智能体使用基于注意力的SE（3）-等变图神经网络，基于其自身的内部状态和从其他智能体接收的信号来更新其位置和内部状态。为了训练这个系统，我们引入了一个新的形状匹配损失的基础上的3D泽尼克多项式，比较预测和目标形状作为连续的空间分布，而不是离散点云，是不变的代理订购，代理数量，和刚体变换。为了实现完全的SO（3）不变性--对旋转不变但对反射敏感，我们包括一个对齐步骤，在计算损失之前，最佳地旋转预测的Zernike光谱以匹配目标。这导致了一个双层问题，其中内部循环优化单位四元数以获得最佳对齐，外部循环更新代理模型。我们使用隐式微分通过对齐步骤计算梯度。我们进行系统的基准测试，以建立我们的形状匹配损失的优势，比其他标准的距离度量形状比较任务。然后，我们证明，AntoMorph可以形成一系列的形状-从简单的椭圆形到复杂的形态-只使用最小的空间线索。
摘要：Biological systems can form complex three-dimensional structures through the collective behavior of identical agents -- cells that follow the same internal rules and communicate without central control. How such distributed control gives rise to precise global patterns remains a central question not only in developmental biology but also in distributed robotics, programmable matter, and multi-agent learning. Here, we introduce DiffeoMorph, an end-to-end differentiable framework for learning a morphogenesis protocol that guides a population of agents to morph into a target 3D shape. Each agent updates its position and internal state using an attention-based SE(3)-equivariant graph neural network, based on its own internal state and signals received from other agents. To train this system, we introduce a new shape-matching loss based on the 3D Zernike polynomials, which compares the predicted and target shapes as continuous spatial distributions, not as discrete point clouds, and is invariant to agent ordering, number of agents, and rigid-body transformations. To enforce full SO(3) invariance -- invariant to rotations yet sensitive to reflections, we include an alignment step that optimally rotates the predicted Zernike spectrum to match the target before computing the loss. This results in a bilevel problem, with the inner loop optimizing a unit quaternion for the best alignment and the outer loop updating the agent model. We compute gradients through the alignment step using implicit differentiation. We perform systematic benchmarking to establish the advantages of our shape-matching loss over other standard distance metrics for shape comparison tasks. We then demonstrate that DiffeoMorph can form a range of shapes -- from simple ellipsoids to complex morphologies -- using only minimal spatial cues.
优化|敛散性(6篇)
【1】Convergence Guarantees for Federated SARSA with Local Training and Heterogeneous Agents
标题：具有本地训练和异类代理的联邦SARSA的融合保证
链接：https://arxiv.org/abs/2512.17688
摘要：我们提出了一种新的理论分析联邦SARSA（FedSARSA）的线性函数逼近和本地训练。我们建立收敛保证FedSARSA在存在异质性，无论是在本地的过渡和奖励，提供第一个样本和通信复杂性的界限，在这种情况下。在我们的分析的核心是一个新的，准确的多步误差扩展单代理SARSA，这是独立的利益。我们的分析精确地量化了异质性的影响，证明了FedSARSA与多个本地更新的收敛性。至关重要的是，我们表明，FedSARSA实现了线性加速相对于代理商的数量，由于马尔可夫抽样的高阶项。数值实验支持我们的理论研究结果。
摘要：We present a novel theoretical analysis of Federated SARSA (FedSARSA) with linear function approximation and local training. We establish convergence guarantees for FedSARSA in the presence of heterogeneity, both in local transitions and rewards, providing the first sample and communication complexity bounds in this setting. At the core of our analysis is a new, exact multi-step error expansion for single-agent SARSA, which is of independent interest. Our analysis precisely quantifies the impact of heterogeneity, demonstrating the convergence of FedSARSA with multiple local updates. Crucially, we show that FedSARSA achieves linear speed-up with respect to the number of agents, up to higher-order terms due to Markovian sampling. Numerical experiments support our theoretical findings.
【2】SCOPE: Sequential Causal Optimization of Process Interventions
标题：范围：过程干预的顺序因果优化
链接：https://arxiv.org/abs/2512.17629
摘要：规范性流程监控（PresPM）建议在业务流程中进行干预，以优化关键绩效指标（KPI）。在现实环境中，干预措施很少是孤立的：组织需要调整干预措施的顺序，以共同指导案件的结果。现有的PresPM方法在这方面存在不足。许多人专注于单一干预决策，而其他人则独立对待多种干预措施，忽略了它们如何随着时间的推移而相互作用。解决这些依赖关系的方法依赖于模拟或数据增强来近似过程以训练强化学习（RL）代理，这可能会造成现实差距并引入偏差。我们引入SCOPE，一种PresPM方法，学习对齐的顺序干预建议。SCOPE采用逆向归纳法来估计每个候选干预行动的效果，将其影响从最终决策点传播回第一个决策点。通过利用因果学习器，我们的方法可以直接利用观察数据，而不像那些需要为强化学习构建过程近似的方法。在现有的合成数据集和新的半合成数据集上的实验表明，SCOPE在优化KPI方面始终优于最先进的PresPM技术。新的半合成设置，基于现实生活中的事件日志，提供作为一个可重用的基准，为今后的工作顺序PresPM。
摘要 ：Prescriptive Process Monitoring (PresPM) recommends interventions during business processes to optimize key performance indicators (KPIs). In realistic settings, interventions are rarely isolated: organizations need to align sequences of interventions to jointly steer the outcome of a case. Existing PresPM approaches fall short in this respect. Many focus on a single intervention decision, while others treat multiple interventions independently, ignoring how they interact over time. Methods that do address these dependencies depend either on simulation or data augmentation to approximate the process to train a Reinforcement Learning (RL) agent, which can create a reality gap and introduce bias. We introduce SCOPE, a PresPM approach that learns aligned sequential intervention recommendations. SCOPE employs backward induction to estimate the effect of each candidate intervention action, propagating its impact from the final decision point back to the first. By leveraging causal learners, our method can utilize observational data directly, unlike methods that require constructing process approximations for reinforcement learning. Experiments on both an existing synthetic dataset and a new semi-synthetic dataset show that SCOPE consistently outperforms state-of-the-art PresPM techniques in optimizing the KPI. The novel semi-synthetic setup, based on a real-life event log, is provided as a reusable benchmark for future work on sequential PresPM.
【3】A Systems-Theoretic View on the Convergence of Algorithms under Disturbances
标题：扰动下算法收敛性的系统论观点
链接：https://arxiv.org/abs/2512.17598
摘要：算法越来越多地在复杂的物理、社会和工程系统中运行，在这些系统中，它们暴露于干扰、噪声以及与其他动态系统的相互连接。这篇文章扩展了已知的孤立操作算法的收敛保证（即，没有干扰），并系统地推导出在这种干扰的存在下的稳定性界限和收敛速度。通过利用逆李雅普诺夫定理，我们推导出关键的不等式，量化干扰的影响。我们进一步展示了如何利用我们的结果来评估各种应用中干扰对算法性能的影响，包括分布式学习中的通信约束，机器学习泛化的敏感性以及隐私的故意噪声注入。这巩固了我们的结果作为一个统一的工具，算法分析中存在的噪声，干扰和与其他动力系统的互连的作用。
摘要：Algorithms increasingly operate within complex physical, social, and engineering systems where they are exposed to disturbances, noise, and interconnections with other dynamical systems. This article extends known convergence guarantees of an algorithm operating in isolation (i.e., without disturbances) and systematically derives stability bounds and convergence rates in the presence of such disturbances. By leveraging converse Lyapunov theorems, we derive key inequalities that quantify the impact of disturbances. We further demonstrate how our result can be utilized to assess the effects of disturbances on algorithmic performance in a wide variety of applications, including communication constraints in distributed learning, sensitivity in machine learning generalization, and intentional noise injection for privacy. This underpins the role of our result as a unifying tool for algorithm analysis in the presence of noise, disturbances, and interconnections with other dynamical systems.
【4】NetworkFF: Unified Layer Optimization in Forward-Only Neural Networks
标题：Networks FF：纯前向神经网络中的统一层优化
链接：https://arxiv.org/abs/2512.17531
备注：Conference paper, IEEE, 2025
摘要：Forward-Forward算法通过正负数据的双重前向传递消除了反向传播的内存限制和生物学上的不可信性。然而，传统的实现遭受关键的层间隔离，其中层独立地优化优度函数，而不利用集体学习动态。这种隔离限制了代表性协调，并限制了更深层次架构中的收敛效率。本文介绍了协作前向-前向（CFF）学习，通过层间合作机制扩展了原始算法，该机制在实现全局上下文集成的同时保留了仅向前计算。我们的框架实现了两种协作范式：具有恒定层间耦合的固定CFF（F-CFF）和具有可学习协作参数的自适应CFF（A-CFF），这些参数在训练过程中不断变化。协作优度函数结合了所有层的加权贡献，在保持记忆效率和生物相容性的同时实现协调的特征学习。对MNIST和Fashion-MNIST的综合评估表明，与基线Forward-Forward实现相比，性能有了显着提高。这些发现建立了层间协作作为Forward-Forward学习的根本增强，并立即适用于神经形态计算架构和能量受限的AI系统。
摘要：The Forward-Forward algorithm eliminates backpropagation's memory constraints and biological implausibility through dual forward passes with positive and negative data. However, conventional implementations suffer from critical inter-layer isolation, where layers optimize goodness functions independently without leveraging collective learning dynamics. This isolation constrains representational coordination and limits convergence efficiency in deeper architectures. This paper introduces Collaborative Forward-Forward (CFF) learning, extending the original algorithm through inter-layer cooperation mechanisms that preserve forward-only computation while enabling global context integration. Our framework implements two collaborative paradigms: Fixed CFF (F-CFF) with constant inter-layer coupling and Adaptive CFF (A-CFF) with learnable collaboration parameters that evolve during training. The collaborative goodness function incorporates weighted contributions from all layers, enabling coordinated feature learning while maintaining memory efficiency and biological plausibility. Comprehensive evaluation on MNIST and Fashion-MNIST demonstrates significant performance improvements over baseline Forward-Forward implementations. These findings establish inter-layer collaboration as a fundamental enhancement to Forward-Forward learning, with immediate applicability to neuromorphic computing architectures and energy-constrained AI systems.
【5】Bridging Training and Merging Through Momentum-Aware Optimization
标题：通过动量感知优化连接训练和合并
链接：https://arxiv.org/abs/2512.17109
备注：Proper is work in progress
摘要：训练大型神经网络和合并特定任务的模型都利用低秩结构，并需要参数重要性估计，但这些挑战一直是孤立的。目前的工作流程在训练过程中计算曲率信息，丢弃它，然后重新计算相似的信息进行合并-浪费计算并丢弃有价值的轨迹数据。我们引入了一个统一的框架，在训练过程中维护因子化的动量和曲率统计，然后重用这些信息进行几何感知模型组合。所提出的方法实现了内存效率相媲美的国家的最先进的方法，同时积累的任务显着性分数，使曲率感知合并没有事后Fisher计算。我们建立了收敛保证的非凸目标的梯度奇异值衰减有界的逼近误差。在自然语言理解基准测试中，曲率感知参数选择在所有稀疏级别上都优于仅幅度基线，多任务合并在强基线上有所改善。与现有的低秩优化器相比，该框架具有秩不变收敛性和优越的超参数鲁棒性。通过将优化轨迹视为可重用资产而不是丢弃它，我们的方法消除了冗余计算，同时实现了更有原则的模型组合。
摘要：Training large neural networks and merging task-specific models both exploit low-rank structure and require parameter importance estimation, yet these challenges have been pursued in isolation. Current workflows compute curvature information during training, discard it, then recompute similar information for merging -- wasting computation and discarding valuable trajectory data. We introduce a unified framework that maintains factorized momentum and curvature statistics during training, then reuses this information for geometry-aware model composition. The proposed method achieves memory efficiency comparable to state-of-the-art approaches while accumulating task saliency scores that enable curvature-aware merging without post-hoc Fisher computation. We establish convergence guarantees for non-convex objectives with approximation error bounded by gradient singular value decay. On natural language understanding benchmarks, curvature-aware parameter selection outperforms magnitude-only baselines across all sparsity levels, with multi-task merging improving over strong baselines. The proposed framework exhibits rank-invariant convergence and superior hyperparameter robustness compared to existing low-rank optimizers. By treating the optimization trajectory as a reusable asset rather than discarding it, our approach eliminates redundant computation while enabling more principled model composition.
【6】Generative Multi-Objective Bayesian Optimization with Scalable Batch Evaluations for Sample-Efficient De Novo Molecular Design
标题：具有可扩展批量评估的生成性多目标Bayesian优化，用于样本高效的从头开始分子设计
链接：https://arxiv.org/abs/2512.17659
摘要：设计必须满足多个，往往相互冲突的目标的分子是分子发现的核心挑战。化学空间的巨大规模和高保真模拟的成本推动了机器学习指导策略的发展，以加速有限数据的设计。其中，贝叶斯优化（BO）为样本高效搜索提供了一个原则性的框架，而生成模型提供了一种机制，可以在固定库之外提出新颖的、多样化的候选人。然而，现有的方法，耦合这两个往往依赖于连续的潜在空间，这引入了架构纠缠和可扩展性的挑战。这项工作介绍了一种替代的，模块化的“生成，然后优化”从头多目标分子设计/发现的框架。在每次迭代中，生成模型用于构建一个大的，不同的候选分子池，之后，一种新的采集函数，qPMHI（多点最大超体积改进概率），用于最佳地选择一批最有可能诱导最大帕累托前沿扩展的候选分子。关键的见解是，qPMHI分解相加，使准确的，可扩展的批次选择，通过简单的排名的概率，可以很容易地估计与蒙特卡洛抽样。我们对最先进的潜在空间和离散分子优化方法的框架进行了基准测试，证明了合成基准测试和应用驱动任务的显着改进。具体而言，在与可持续能源存储相关的案例研究中，我们表明，我们的方法可以快速发现用于水性氧化还原液流电池应用的新型，多样化和高性能有机（醌基）阴极材料。
摘要：Designing molecules that must satisfy multiple, often conflicting objectives is a central challenge in molecular discovery. The enormous size of chemical space and the cost of high-fidelity simulations have driven the development of machine learning-guided strategies for accelerating design with limited data. Among these, Bayesian optimization (BO) offers a principled framework for sample-efficient search, while generative models provide a mechanism to propose novel, diverse candidates beyond fixed libraries. However, existing methods that couple the two often rely on continuous latent spaces, which introduces both architectural entanglement and scalability challenges. This work introduces an alternative, modular "generate-then-optimize" framework for de novo multi-objective molecular design/discovery. At each iteration, a generative model is used to construct a large, diverse pool of candidate molecules, after which a novel acquisition function, qPMHI (multi-point Probability of Maximum Hypervolume Improvement), is used to optimally select a batch of candidates most likely to induce the largest Pareto front expansion. The key insight is that qPMHI decomposes additively, enabling exact, scalable batch selection via only simple ranking of probabilities that can be easily estimated with Monte Carlo sampling. We benchmark the framework against state-of-the-art latent-space and discrete molecular optimization methods, demonstrating significant improvements across synthetic benchmarks and application-driven tasks. Specifically, in a case study related to sustainable energy storage, we show that our approach quickly uncovers novel, diverse, and high-performing organic (quinone-based) cathode materials for aqueous redox flow battery applications.
预测|估计(3篇)
【1】When Data Quality Issues Collide: A Large-Scale Empirical Study of Co-Occurring Data Quality Issues in Software Defect Prediction
标题：当数据质量问题碰撞时：软件缺陷预测中共存数据质量问题的大规模实证研究
链接：https://arxiv.org/abs/2512.17460
摘要：软件缺陷预测（SDP）模型是主动软件质量保证的核心，但其有效性往往受到可用数据集质量的限制。以前的研究通常只研究单个问题，如类不平衡或功能无关性，而忽略了现实世界中的数据问题经常同时发生并相互作用。据我们所知，这项研究提出了SDP中的第一个大规模实证分析，该分析同时检查了374个数据集和5个分类器中的5个共存数据质量问题（类不平衡，类重叠，不相关特征，属性噪声和离群值）。我们采用可解释的提升机与分层交互分析来量化默认超参数设置下的直接和条件效应，反映实际的基线使用。 我们的研究结果表明，共现几乎是普遍的：即使是最不常见的问题（属性噪声）也会在超过93%的数据集中与其他问题一起出现。不相关的特征和不平衡几乎无处不在，而类重叠是最常见的有害问题。我们确定了类重叠的稳定临界点为0.20，不平衡的临界点为0.65-0.70，不相关的临界点为0.94，超过这个临界点，大多数模型开始退化。我们还发现了违反直觉的模式，例如当不相关的特征较低时，离群值可以提高性能，这强调了上下文感知评估的重要性。最后，我们揭示了性能鲁棒性的权衡：没有一个学习者在所有条件下都占主导地位。 通过共同分析患病率，共现，阈值和条件效应，我们的研究直接解决了SDP研究中的一个持续存在的差距。因此，超越孤立的分析，提供全面的，数据感知的理解，质量问题如何在现实世界中塑造模型性能。
摘要：Software Defect Prediction (SDP) models are central to proactive software quality assurance, yet their effectiveness is often constrained by the quality of available datasets. Prior research has typically examined single issues such as class imbalance or feature irrelevance in isolation, overlooking that real-world data problems frequently co-occur and interact. This study presents, to our knowledge, the first large-scale empirical analysis in SDP that simultaneously examines five co-occurring data quality issues (class imbalance, class overlap, irrelevant features, attribute noise, and outliers) across 374 datasets and five classifiers. We employ Explainable Boosting Machines together with stratified interaction analysis to quantify both direct and conditional effects under default hyperparameter settings, reflecting practical baseline usage. Our results show that co-occurrence is nearly universal: even the least frequent issue (attribute noise) appears alongside others in more than 93% of datasets. Irrelevant features and imbalance are nearly ubiquitous, while class overlap is the most consistently harmful issue. We identify stable tipping points around 0.20 for class overlap, 0.65-0.70 for imbalance, and 0.94 for irrelevance, beyond which most models begin to degrade. We also uncover counterintuitive patterns, such as outliers improving performance when irrelevant features are low, underscoring the importance of context-aware evaluation. Finally, we expose a performance-robustness trade-off: no single learner dominates under all conditions. By jointly analyzing prevalence, co-occurrence, thresholds, and conditional effects, our study directly addresses a persistent gap in SDP research. Hence, moving beyond isolated analyses to provide a holistic, data-aware understanding of how quality issues shape model performance in real-world settings.
【2】Physics-Informed Lightweight Machine Learning for Aviation Visibility Nowcasting Across Multiple Climatic Regimes
标题：基于物理知识的轻量级机器学习，用于跨多个气候条件的航空可见性现播
链接：https://arxiv.org/abs/2512.16967
备注：12 pages, 5 tables, 1 figure. Uses publicly available METAR surface observations and TAF forecast data for benchmarking
摘要：低能见度和降水事件的短期预报（临近预报）对于航空安全和运营效率至关重要。目前的业务方法依赖于计算密集型数值天气预报指导和人类发布的TAF产品，这些产品通常表现出保守的偏差和有限的时间分辨率。这项研究提出了一个轻量级的梯度提升框架（XGBoost），专门针对表面观测数据（METAR）进行训练，并通过基于热力学原理的物理指导特征工程进行增强。该框架使用2000年至2024年的历史数据对代表不同气候状况的11个国际机场（包括SCEL，KJFK，KORD，KDEN，SBGR和VIDP）进行了评估。结果表明，该模型成功地捕捉到当地的物理过程，而无需手动配置。在对作战TAF预测的盲比较评估中，自动化模型在战术范围（3小时）内实现了更高的检测率，召回率提高了2.5至4.0倍，同时减少了误报。此外，SHAP分析表明，该模型进行了隐式重建当地的物理驱动程序（平流，辐射和沉降），提供可操作的可解释性的操作态势感知。 保留字：航空气象学物理学引导的机器学习;可解释的人工智能;轻量级机器学习;临近预报; METAR; TAF验证;边缘计算
摘要：Short-term prediction (nowcasting) of low-visibility and precipitation events is critical for aviation safety and operational efficiency. Current operational approaches rely on computationally intensive numerical weather prediction guidance and human-issued TAF products, which often exhibit conservative biases and limited temporal resolution. This study presents a lightweight gradient boosting framework (XGBoost) trained exclusively on surface observation data (METAR) and enhanced through physics-guided feature engineering based on thermodynamic principles. The framework is evaluated across 11 international airports representing distinct climatic regimes (including SCEL, KJFK, KORD, KDEN, SBGR, and VIDP) using historical data from 2000 to 2024. Results suggest that the model successfully captures underlying local physical processes without manual configuration. In a blind comparative evaluation against operational TAF forecasts, the automated model achieved substantially higher detection rates at tactical horizons (3 hours), with a 2.5 to 4.0 times improvement in recall while reducing false alarms. Furthermore, SHAP analysis reveals that the model performs an implicit reconstruction of local physical drivers (advection, radiation, and subsidence), providing actionable explainability for operational situational awareness. Keywords: aviation meteorology; physics-guided machine learning; explainable artificial intelligence; lightweight machine learning; nowcasting; METAR; TAF verification; edge computing
【3】Sharp Structure-Agnostic Lower Bounds for General Functional Estimation
标题：一般功能估计的尖锐结构不可知下限
链接：https://arxiv.org/abs/2512.17341
备注：95 pages; generalize and subsume partial results of arXiv:2402.14264 by the same authors
摘要：设计有效的非参数估计量一直是统计学、机器学习和决策中的一个中心问题。经典的优化过程通常依赖于强结构性假设，这些假设在实践中可能会被错误指定，并使部署复杂化。这一限制引发了人们对结构不可知方法的兴趣，这种方法可以消除黑盒干扰估计的偏见，而不会强加结构先验。因此，了解这些方法的基本局限性至关重要。本文提供了一个系统的调查，可实现的结构不可知估计的最佳错误率。我们首先表明，估计平均治疗效果（ATE），在因果推理的中心参数，双鲁棒学习达到最佳的结构不可知的错误率。然后，我们将我们的分析扩展到依赖于未知滋扰函数的一般类泛函，并建立去偏/双机器学习（DML）的结构不可知最优性。我们区分两个制度-一个双鲁棒性是可以实现的，一个是不-导致不同的最佳速率为一阶去偏，并表明DML是最佳的，在这两个制度。最后，我们实例化我们的一般下界，得到明确的最优利率，恢复现有的结果，并扩展到额外的被估量的兴趣。我们的研究结果为广泛使用的一阶去偏方法提供了理论验证，并为在没有结构假设的情况下寻求最佳方法的从业者提供了指导。本文推广并包含了作者在[jin 2024 structure]中建立的ATE下界.
摘要：The design of efficient nonparametric estimators has long been a central problem in statistics, machine learning, and decision making. Classical optimal procedures often rely on strong structural assumptions, which can be misspecified in practice and complicate deployment. This limitation has sparked growing interest in structure-agnostic approaches -- methods that debias black-box nuisance estimates without imposing structural priors. Understanding the fundamental limits of these methods is therefore crucial. This paper provides a systematic investigation of the optimal error rates achievable by structure-agnostic estimators. We first show that, for estimating the average treatment effect (ATE), a central parameter in causal inference, doubly robust learning attains optimal structure-agnostic error rates. We then extend our analysis to a general class of functionals that depend on unknown nuisance functions and establish the structure-agnostic optimality of debiased/double machine learning (DML). We distinguish two regimes -- one where double robustness is attainable and one where it is not -- leading to different optimal rates for first-order debiasing, and show that DML is optimal in both regimes. Finally, we instantiate our general lower bounds by deriving explicit optimal rates that recover existing results and extend to additional estimands of interest. Our results provide theoretical validation for widely used first-order debiasing methods and guidance for practitioners seeking optimal approaches in the absence of structural assumptions. This paper generalizes and subsumes the ATE lower bound established in \citet{jin2024structure} by the same authors.
其他神经网络|深度学习|模型|建模(19篇)
【1】Distributionally Robust Imitation Learning: Layered Control Architecture for Certifiable Autonomy
标题：分布式稳健模仿学习：可认证自主性的分层控制架构
链接：https://arxiv.org/abs/2512.17899
备注：18 pages, 5 figures
摘要：模仿学习（IL）通过从专家演示中学习来实现自主行为。虽然IL比强化学习等比较替代方案更有效，但它对分布变化引起的复合误差很敏感。在系统上使用基于IL的反馈律时，有两个重要的分布偏移来源：由政策错误引起的分布偏移和由于外部干扰和由于缺乏学习引起的内生模型错误引起的分布偏移。我们以前开发的方法，泰勒级数模仿学习（TaSIL）和$\mathcal{L}_1$ -分布鲁棒自适应控制（\ellonedrac），解决了互补的方式分布变化的挑战。虽然TaSIL提供了对政策错误引起的分布变化的鲁棒性，但\ellonedrac提供了对由于任意和认知不确定性引起的分布变化的鲁棒性。为了使可认证的IL学习和/或不确定的动态系统，我们制定了{分布式鲁棒模仿策略（DRIP）}架构，一个分层控制架构（LCA），集成TaSIL和~\ellonedrac。通过明智地设计以层为中心的输入和输出要求，我们展示了如何保证整个控制管道的证书。我们的解决方案为设计完全可认证的自主管道铺平了道路，通过建议的LCA方法将基于学习的组件（如感知）与可认证的基于模型的决策相集成。
摘要：Imitation learning (IL) enables autonomous behavior by learning from expert demonstrations. While more sample-efficient than comparative alternatives like reinforcement learning, IL is sensitive to compounding errors induced by distribution shifts. There are two significant sources of distribution shifts when using IL-based feedback laws on systems: distribution shifts caused by policy error and distribution shifts due to exogenous disturbances and endogenous model errors due to lack of learning. Our previously developed approaches, Taylor Series Imitation Learning (TaSIL) and $\mathcal{L}_1$ -Distributionally Robust Adaptive Control (\ellonedrac), address the challenge of distribution shifts in complementary ways. While TaSIL offers robustness against policy error-induced distribution shifts, \ellonedrac offers robustness against distribution shifts due to aleatoric and epistemic uncertainties. To enable certifiable IL for learned and/or uncertain dynamical systems, we formulate \textit{Distributionally Robust Imitation Policy (DRIP)} architecture, a Layered Control Architecture (LCA) that integrates TaSIL and~\ellonedrac. By judiciously designing individual layer-centric input and output requirements, we show how we can guarantee certificates for the entire control pipeline. Our solution paves the path for designing fully certifiable autonomy pipelines, by integrating learning-based components, such as perception, with certifiable model-based decision-making through the proposed LCA approach.
【2】Regularized Random Fourier Features and Finite Element Reconstruction for Operator Learning in Sobolev Space
标题：Sobolev空间中算子学习的正则化随机Fourier特征和有限元重构
链接：https://arxiv.org/abs/2512.17884
摘要：算子学习是无限维函数空间之间映射的数据驱动近似，例如偏微分方程的解算子。基于核的算子学习可以提供准确的、理论上合理的近似，比标准方法需要更少的训练。然而，它们对于大型训练集可能变得计算上难以承受，并且可能对噪声敏感。我们提出了一个正则化的随机傅立叶特征（RRFF）的方法，再加上有限元重建映射（RRFF-FEM），学习运营商从嘈杂的数据。该方法使用从多变量学生的$t$分布，以及频率加权Tikhonov正则化，抑制高频噪声的随机功能。我们建立了高概率界上的极端奇异值相关联的随机特征矩阵，并表明，当功能$N$规模像$m \log m$与训练样本的数量$m$，系统是良好的条件下，产生估计和推广的保证。对基准PDE问题（包括平流、Burgers、Darcy流、Helmholtz、Navier-Stokes和结构力学）的详细数值实验表明，与非正则化随机特征模型相比，RRFF和RRFF-FEM对噪声具有鲁棒性，并且在减少训练时间的情况下实现了改进的性能，同时相对于内核和神经操作符测试保持了有竞争力的精度。
摘要：Operator learning is a data-driven approximation of mappings between infinite-dimensional function spaces, such as the solution operators of partial differential equations. Kernel-based operator learning can offer accurate, theoretically justified approximations that require less training than standard methods. However, they can become computationally prohibitive for large training sets and can be sensitive to noise. We propose a regularized random Fourier feature (RRFF) approach, coupled with a finite element reconstruction map (RRFF-FEM), for learning operators from noisy data. The method uses random features drawn from multivariate Student's $t$ distributions, together with frequency-weighted Tikhonov regularization that suppresses high-frequency noise. We establish high-probability bounds on the extreme singular values of the associated random feature matrix and show that when the number of features $N$ scales like $m \log m$ with the number of training samples $m$, the system is well-conditioned, which yields estimation and generalization guarantees. Detailed numerical experiments on benchmark PDE problems, including advection, Burgers', Darcy flow, Helmholtz, Navier-Stokes, and structural mechanics, demonstrate that RRFF and RRFF-FEM are robust to noise and achieve improved performance with reduced training time compared to the unregularized random feature model, while maintaining competitive accuracy relative to kernel and neural operator tests.
【3】Calibratable Disambiguation Loss for Multi-Instance Partial-Label Learning
标题：多实例部分标签学习的可校准歧义消除损失
链接：https://arxiv.org/abs/2512.17788
摘要：多实例部分标签学习（MIPL）是一种弱监督框架，它扩展了多实例学习（MIL）和部分标签学习（PLL）的原理，以解决实例和标签空间中不精确监督的挑战。然而，现有的MIPL方法经常遭受不良校准，破坏分类器的可靠性。在这项工作中，我们提出了一个即插即用的可校准的消歧损失（CDL），同时提高分类精度和校准性能。损失有两个实例：第一个基于候选标签集的概率校准预测，而第二个集成了候选和非候选标签集的概率。建议的CDL可以无缝地纳入现有的MIPL和PLL框架。我们提供了一个理论分析，建立了下限和正则化性质的CDL，证明了其优越性，比传统的消歧损失。在基准数据集和真实数据集上的实验结果证实，我们的CDL显著提高了分类和校准性能。
摘要：Multi-instance partial-label learning (MIPL) is a weakly supervised framework that extends the principles of multi-instance learning (MIL) and partial-label learning (PLL) to address the challenges of inexact supervision in both instance and label spaces. However, existing MIPL approaches often suffer from poor calibration, undermining classifier reliability. In this work, we propose a plug-and-play calibratable disambiguation loss (CDL) that simultaneously improves classification accuracy and calibration performance. The loss has two instantiations: the first one calibrates predictions based on probabilities from the candidate label set, while the second one integrates probabilities from both candidate and non-candidate label sets. The proposed CDL can be seamlessly incorporated into existing MIPL and PLL frameworks. We provide a theoretical analysis that establishes the lower bound and regularization properties of CDL, demonstrating its superiority over conventional disambiguation losses. Experimental results on benchmark and real-world datasets confirm that our CDL significantly enhances both classification and calibration performance.
【4】Vidarc: Embodied Video Diffusion Model for Closed-loop Control
标题：Vidarc：用于闭环控制的同步视频扩散模型
链接：https://arxiv.org/abs/2512.17661
摘要：在数据稀缺的环境中，由于复杂的实施动态和不同的上下文，机器人手臂操纵是一项极具挑战性的任务。最近的基于视频的方法已经显示出很大的希望，在捕捉和传输的时间和物理的相互作用，通过预训练的互联网规模的视频数据。然而，这样的方法通常没有针对特定于测量的闭环控制进行优化，通常遭受高延迟和不充分的接地。在本文中，我们提出了Vidarc（视频扩散动作推理和闭环控制），一种新的自回归体现视频扩散方法增强了一个掩蔽的逆动力学模型。通过将视频预测与动作相关的掩码结合起来，并通过缓存的自回归生成来整合实时反馈，Vidarc实现了快速、准确的闭环控制。Vidarc经过100万个跨具体化集的预训练，超越了最先进的基线，在现实世界的部署中实现了至少15%的成功率和91%的延迟减少。我们还强调了它在以前看不见的机器人平台上强大的泛化和纠错能力。
摘要：Robotic arm manipulation in data-scarce settings is a highly challenging task due to the complex embodiment dynamics and diverse contexts. Recent video-based approaches have shown great promise in capturing and transferring the temporal and physical interactions by pre-training on Internet-scale video data. However, such methods are often not optimized for the embodiment-specific closed-loop control, typically suffering from high latency and insufficient grounding. In this paper, we present Vidarc (Video Diffusion for Action Reasoning and Closed-loop Control), a novel autoregressive embodied video diffusion approach augmented by a masked inverse dynamics model. By grounding video predictions with action-relevant masks and incorporating real-time feedback through cached autoregressive generation, Vidarc achieves fast, accurate closed-loop control. Pre-trained on one million cross-embodiment episodes, Vidarc surpasses state-of-the-art baselines, achieving at least a 15% higher success rate in real-world deployment and a 91% reduction in latency. We also highlight its robust generalization and error correction capabilities across previously unseen robotic platforms.
【5】Estimating Spatially Resolved Radiation Fields Using Neural Networks
标题：使用神经网络估计空间分辨辐射场
链接：https://arxiv.org/abs/2512.17654
摘要：我们提出了一个深入的分析如何建立和训练神经网络，以估计散射辐射场的空间分布的辐射防护剂量学在医疗辐射领域，如那些发现在介入放射学和心脏病学。因此，我们提出了三个不同的综合生成的数据集，训练的复杂性越来越高，使用基于Geant 4的Monte-Carlo模拟应用程序。在这些数据集上，我们评估了神经网络的卷积和全连接架构，以证明哪些设计决策可以很好地重建此类辐射场空间域上的注量和光谱分布。所有使用的数据集以及我们的训练管道都在单独的存储库中作为开源发布。
摘要：We present an in-depth analysis on how to build and train neural networks to estimate the spatial distribution of scattered radiation fields for radiation protection dosimetry in medical radiation fields, such as those found in Interventional Radiology and Cardiology. Therefore, we present three different synthetically generated datasets with increasing complexity for training, using a Monte-Carlo Simulation application based on Geant4. On those datasets, we evaluate convolutional and fully connected architectures of neural networks to demonstrate which design decisions work well for reconstructing the fluence and spectra distributions over the spatial domain of such radiation fields. All used datasets as well as our training pipeline are published as open source in separate repositories.
【6】Sharing Knowledge without Sharing Data: Stitches can improve ensembles of disjointly trained models
标题：在不共享数据的情况下共享知识：缝合可以改进不连续训练的模型的集合
链接：https://arxiv.org/abs/2512.17592
备注：35 pages, 11 figures
摘要：深度学习已经被证明非常有能力执行许多现实世界的任务。然而，这种性能通常取决于大型和各种数据集的存在。在某些情况下，例如在医疗领域，数据往往分散在各方之间，并且无法轻易共享。虽然联邦学习解决了这种情况，但它是一种解决方案，需要各方同步训练单个模型，交换有关模型权重的信息。我们研究了异步协作如何影响性能，其中只共享已经训练好的模型（例如，作为发布的一部分），并建议使用拼接作为组合模型的方法。 通过采取多目标的视角，其中独立地查看各方数据的性能，我们发现，当考虑对该单一方数据的性能时，仅对单一方数据进行训练，当与另一方数据合并时，会导致类似的性能，而对其他方数据的性能则明显更差。此外，虽然这种单独训练的网络的集合可以更好地推广，但各方自己的数据集的性能会受到影响。我们发现，将单独训练的模型中的中间表示与一对位置良好的缝合层相结合，可以使这种性能恢复到具有竞争力的程度，同时保持改进的泛化能力，这表明异步协作可以产生有竞争力的结果。
摘要 ：Deep learning has been shown to be very capable at performing many real-world tasks. However, this performance is often dependent on the presence of large and varied datasets. In some settings, like in the medical domain, data is often fragmented across parties, and cannot be readily shared. While federated learning addresses this situation, it is a solution that requires synchronicity of parties training a single model together, exchanging information about model weights. We investigate how asynchronous collaboration, where only already trained models are shared (e.g. as part of a publication), affects performance, and propose to use stitching as a method for combining models. Through taking a multi-objective perspective, where performance on each parties' data is viewed independently, we find that training solely on a single parties' data results in similar performance when merging with another parties' data, when considering performance on that single parties' data, while performance on other parties' data is notably worse. Moreover, while an ensemble of such individually trained networks generalizes better, performance on each parties' own dataset suffers. We find that combining intermediate representations in individually trained models with a well placed pair of stitching layers allows this performance to recover to a competitive degree while maintaining improved generalization, showing that asynchronous collaboration can yield competitive results.
【7】Deep Learning-Based Surrogate Creep Modelling in Inconel 625: A High-Temperature Alloy Study
标题：基于深度学习的Inconel 625替代品蠕动建模：高温合金研究
链接：https://arxiv.org/abs/2512.17477
备注：Presented in 10th International Congress on Computational Mechanics and Simulation (ICCMS) 2025, IIT Bhubaneswar
摘要：在Inconel 625等高温合金中，随时间变化的变形，特别是蠕变，是航空航天和能源系统中使用的部件长期可靠性的关键因素。虽然Inconel 625显示出优异的抗蠕变性，但ANSYS等工具中的有限元蠕变模拟仍然计算昂贵，通常需要数十分钟才能完成一次10，000小时的运行。这项工作提出了基于深度学习的代理模型，为此类模拟提供快速准确的替代。蠕变应变数据是在ANSYS中使用诺顿定律在50至150 MPa的单轴应力和700至1000 $^\circ$C的温度下生成的，并且该时间数据集用于训练两种架构：用于不确定性感知和生成预测的BiLSTM变分自动编码器，以及采用自注意力捕获长期时间行为的BiLSTM Transformer混合。这两个模型都充当替代预测器，BiLSTM-VAE提供概率输出，BiLSTM-Transformer提供高确定性精度。使用RMSE、MAE和$R^2$评估性能。结果表明，BiLSTM-VAE提供了稳定可靠的蠕变应变预测，而BiLSTM-Transformer在整个时间范围内实现了很高的准确性。延迟测试表明速度大幅提高：对于给定的应力-温度条件，每个ANSYS模拟需要30到40分钟，而代理模型在几秒钟内就可以生成预测。所提出的框架能够快速蠕变评估的设计优化和结构健康监测，并提供了一个可扩展的解决方案，高温合金应用。
摘要：Time-dependent deformation, particularly creep, in high-temperature alloys such as Inconel 625 is a key factor in the long-term reliability of components used in aerospace and energy systems. Although Inconel 625 shows excellent creep resistance, finite-element creep simulations in tools such as ANSYS remain computationally expensive, often requiring tens of minutes for a single 10,000-hour run. This work proposes deep learning based surrogate models to provide fast and accurate replacements for such simulations. Creep strain data was generated in ANSYS using the Norton law under uniaxial stresses of 50 to 150 MPa and temperatures of 700 to 1000 $^\circ$C, and this temporal dataset was used to train two architectures: a BiLSTM Variational Autoencoder for uncertainty-aware and generative predictions, and a BiLSTM Transformer hybrid that employs self-attention to capture long-range temporal behavior. Both models act as surrogate predictors, with the BiLSTM-VAE offering probabilistic output and the BiLSTM-Transformer delivering high deterministic accuracy. Performance is evaluated using RMSE, MAE, and $R^2$. Results show that the BiLSTM-VAE provides stable and reliable creep strain forecasts, while the BiLSTM-Transformer achieves strong accuracy across the full time range. Latency tests indicate substantial speedup: while each ANSYS simulation requires 30 to 40 minutes for a given stress-temperature condition, the surrogate models produce predictions within seconds. The proposed framework enables rapid creep assessment for design optimization and structural health monitoring, and provides a scalable solution for high-temperature alloy applications.
【8】Task Schema and Binding: A Double Dissociation Study of In-Context Learning
标题：任务图式与绑定：情境学习的双重分离研究
链接：https://arxiv.org/abs/2512.17325
备注：20pages, 2figures
摘要：我们提供了因果机制验证，在上下文学习（ICL）分解成两个可分离的机制：任务模式（抽象的任务类型识别）和绑定（具体的输入输出关联）。通过对来自7个Transformer系列的9个模型加上Mamba（370 M-13 B参数）的激活修补实验，我们建立了三个关键发现： 1.双重解离：通过后期MLP修补，任务模式传输率为100%;通过剩余流修补，绑定传输率为62%--证明了可分离的机制 2.先验图式权衡：图式依赖与先验知识呈负相关（Spearman rho =-0.596，p < 0.001，N=28个任务模型对） 3.架构通用性：该机制可在所有测试架构中运行，包括非Transformer Mamba 这些研究结果提供了一个机械的ICL难题，与以前的观点处理ICL作为一个整体的机制（无论是检索为基础，梯度下降，或纯粹的贝叶斯）。通过建立图式和绑定是神经可分离的-不仅仅是行为模式-我们提供了ICL的双过程理论的因果证据。当先验知识不存在时，模型依赖于任务图式，但先验知识通过注意错误路由（72.7%的近因偏差）而不是直接输出竞争（0%）进行干扰。这解释了为什么任意映射成功（零优先级导致完全模式依赖），而事实覆盖失败-并揭示了真正的瓶颈是注意力，而不是输出级别。实际影响：了解这些双重机制可以实现更高效的快速工程-可靠的模式传输减少了新任务所需的演示，而先验感知设计可以降低高先验场景中38%的绑定失败率，提高生产部署中的ICL系统可靠性。
摘要：We provide causal mechanistic validation that in-context learning (ICL) decomposes into two separable mechanisms: Task Schema (abstract task type recognition) and Binding (specific input-output associations). Through activation patching experiments across 9 models from 7 Transformer families plus Mamba (370M-13B parameters), we establish three key findings: 1. Double dissociation: Task Schema transfers at 100% via late MLP patching; Binding transfers at 62% via residual stream patching -- proving separable mechanisms 2. Prior-Schema trade-off: Schema reliance inversely correlates with prior knowledge (Spearman rho = -0.596, p < 0.001, N=28 task-model pairs) 3. Architecture generality: The mechanism operates across all tested architectures including the non-Transformer Mamba These findings offer a mechanistic account of the ICL puzzle that contrasts with prior views treating ICL as a monolithic mechanism (whether retrieval-based, gradient descent-like, or purely Bayesian). By establishing that Schema and Binding are neurally dissociable -- not merely behavioral modes -- we provide causal evidence for dual-process theories of ICL. Models rely on Task Schema when prior knowledge is absent, but prior knowledge interferes through attentional mis-routing (72.7% recency bias) rather than direct output competition (0%). This explains why arbitrary mappings succeed (zero prior leads to full Schema reliance) while factual overrides fail -- and reveals that the true bottleneck is attentional, not output-level. Practical implications: Understanding these dual mechanisms enables more efficient prompt engineering -- reliable schema transfer reduces required demonstrations for novel tasks, while prior-aware design can mitigate the 38% binding failure rate in high-prior scenarios, improving ICL system reliability in production deployments.
【9】M2RU: Memristive Minion Recurrent Unit for Continual Learning at the Edge
标题：M2 RU：边缘持续学习的Memristive Miniion Recurrent Unit
链接：https://arxiv.org/abs/2512.17299
摘要：边缘平台上的持续学习仍然具有挑战性，因为循环网络依赖于能源密集型训练过程和频繁的数据移动，这对于嵌入式部署来说是不切实际的。这项工作介绍了M2 RU，一种混合信号架构，实现了有效的时间处理与片上持续学习的奴才经常性单位。该架构集成了加权位流，使多位数字输入能够在交叉开关中处理，而无需高分辨率转换，并集成了经验重放机制，可在域偏移下稳定学习。M2 RU在48.62 mW下实现15 GOPS，相当于每瓦312 GOPS，并在顺序MNIST和CIFAR-10任务中将精度保持在软件基线的5%以内。与CMOS数字设计相比，该加速器的能效提高了29倍。设备感知分析显示，在持续学习工作负载下，预期使用寿命为12.2年。这些结果建立M2 RU作为一个可扩展的和节能的平台，在边缘级的时间智能实时适应。
摘要 ：Continual learning on edge platforms remains challenging because recurrent networks depend on energy-intensive training procedures and frequent data movement that are impractical for embedded deployments. This work introduces M2RU, a mixed-signal architecture that implements the minion recurrent unit for efficient temporal processing with on-chip continual learning. The architecture integrates weighted-bit streaming, which enables multi-bit digital inputs to be processed in crossbars without high-resolution conversion, and an experience replay mechanism that stabilizes learning under domain shifts. M2RU achieves 15 GOPS at 48.62 mW, corresponding to 312 GOPS per watt, and maintains accuracy within 5 percent of software baselines on sequential MNIST and CIFAR-10 tasks. Compared with a CMOS digital design, the accelerator provides 29X improvement in energy efficiency. Device-aware analysis shows an expected operational lifetime of 12.2 years under continual learning workloads. These results establish M2RU as a scalable and energy-efficient platform for real-time adaptation in edge-level temporal intelligence.
【10】Learning solution operator of dynamical systems with diffusion maps kernel ridge regression
标题：具有扩散映射的动力系统学习解操作核岭回归
链接：https://arxiv.org/abs/2512.17203
摘要：许多科学和工程系统表现出复杂的非线性动力学，很难在长时间内准确预测。虽然数据驱动模型已经显示出了希望，但当控制长期行为的几何结构未知或表现不佳时，它们的性能往往会恶化。我们证明了一个简单的核岭回归（KRR）框架，结合动态感知验证策略，为复杂动态系统的长期预测提供了一个强大的基线。通过采用源自扩散映射的数据驱动内核，所提出的扩散映射内核岭回归（DM-KRR）方法隐式地适应系统不变集的内在几何结构，而不需要显式的流形重建或吸引子建模，这些过程通常会限制预测性能。在广泛的系统中，包括光滑流形，混沌吸引子和高维时空流，DM-KRR在准确性和数据效率方面始终优于最先进的随机特征，神经网络和算子学习方法。这些研究结果强调，长期预测技能不仅取决于模型的表现力，但关键是尊重几何约束编码的数据，通过动态一致的模型选择。简单性、几何意识和强大的经验性能为复杂动力系统的可靠和有效学习指明了一条有前途的道路。
摘要：Many scientific and engineering systems exhibit complex nonlinear dynamics that are difficult to predict accurately over long time horizons. Although data-driven models have shown promise, their performance often deteriorates when the geometric structures governing long-term behavior are unknown or poorly represented. We demonstrate that a simple kernel ridge regression (KRR) framework, when combined with a dynamics-aware validation strategy, provides a strong baseline for long-term prediction of complex dynamical systems. By employing a data-driven kernel derived from diffusion maps, the proposed Diffusion Maps Kernel Ridge Regression (DM-KRR) method implicitly adapts to the intrinsic geometry of the system's invariant set, without requiring explicit manifold reconstruction or attractor modeling, procedures that often limit predictive performance. Across a broad range of systems, including smooth manifolds, chaotic attractors, and high-dimensional spatiotemporal flows, DM-KRR consistently outperforms state-of-the-art random feature, neural-network and operator-learning methods in both accuracy and data efficiency. These findings underscore that long-term predictive skill depends not only on model expressiveness, but critically on respecting the geometric constraints encoded in the data through dynamically consistent model selection. Together, simplicity, geometry awareness, and strong empirical performance point to a promising path for reliable and efficient learning of complex dynamical systems.
【11】BumpNet: A Sparse Neural Network Framework for Learning PDE Solutions
标题：BumpNet：用于学习DTE解决方案的稀疏神经网络框架
链接：https://arxiv.org/abs/2512.17198
摘要：我们介绍BumpNet，一个用于PDE数值解和算子学习的稀疏神经网络框架。BumpNet基于无网格基函数扩展，与径向基函数（RBF）网络类似。与RBF网络不同，BumpNet中的基函数是由普通的sigmoid激活函数构造的。这使得能够有效地使用针对此类网络优化的现代训练技术。基函数的所有参数，包括形状、位置和幅度，都是完全可训练的。通过在训练过程中动态修剪基函数，有效地实现了模型的简约性和h-自适应性。BumpNet是一个通用框架，可以与现有的神经架构相结合，用于学习PDE解决方案：在这里，我们提出了Bump-PINNs（具有物理信息神经网络的BumpNet）用于解决一般PDE; Bump-EDNN（具有进化深度神经网络的BumpNet）用于解决时间演化PDE; Bump-DeepONet（具有深度算子网络的BumpNet）用于PDE算子学习。Bump-PINN使用与PINN相同的基于搭配的方法进行训练，Bump-EDNN仅在空间域中使用BumpNet，并使用EDNN来及时推进解决方案，而Bump-DeepONets使用BumpNet回归网络作为DeepONet的主干网络。大量的数值实验表明，所提出的架构的效率和准确性。
摘要：We introduce BumpNet, a sparse neural network framework for PDE numerical solution and operator learning. BumpNet is based on meshless basis function expansion, in a similar fashion to radial-basis function (RBF) networks. Unlike RBF networks, the basis functions in BumpNet are constructed from ordinary sigmoid activation functions. This enables the efficient use of modern training techniques optimized for such networks. All parameters of the basis functions, including shape, location, and amplitude, are fully trainable. Model parsimony and h-adaptivity are effectively achieved through dynamically pruning basis functions during training. BumpNet is a general framework that can be combined with existing neural architectures for learning PDE solutions: here, we propose Bump-PINNs (BumpNet with physics-informed neural networks) for solving general PDEs; Bump-EDNN (BumpNet with evolutionary deep neural networks) to solve time-evolution PDEs; and Bump-DeepONet (BumpNet with deep operator networks) for PDE operator learning. Bump-PINNs are trained using the same collocation-based approach used by PINNs, Bump-EDNN uses a BumpNet only in the spatial domain and uses EDNNs to advance the solution in time, while Bump-DeepONets employ a BumpNet regression network as the trunk network of a DeepONet. Extensive numerical experiments demonstrate the efficiency and accuracy of the proposed architecture.
【12】Fault Diagnosis and Quantification for Photovoltaic Arrays based on Differentiable Physical Models
标题：基于可区分物理模型的太阳能电池阵故障诊断与量化
链接：https://arxiv.org/abs/2512.17107
摘要：准确的故障诊断和量化对于光伏阵列的可靠运行和智能维护至关重要。然而，现有的故障量化方法往往受到有限的效率和可解释性。为了解决这些问题，本文提出了一种新的光伏串故障量化方法的基础上可微快速故障仿真模型（DFFSM）。建议DFFSM准确地模拟多故障下的I-V特性，并提供故障参数的分析梯度。利用这一属性，基于梯度的故障参数识别（GFPI）的方法，使用Adahessian优化器的开发，以有效地量化部分阴影，短路和串联电阻退化。模拟和测量I-V曲线的实验结果表明，所提出的GFPI在不同故障下均达到了较高的量化精度，I-V重建误差低于3%，证实了可微物理模拟器应用于光伏系统故障诊断的可行性和有效性。
摘要：Accurate fault diagnosis and quantification are essential for the reliable operation and intelligent maintenance of photovoltaic (PV) arrays. However, existing fault quantification methods often suffer from limited efficiency and interpretability. To address these challenges, this paper proposes a novel fault quantification approach for PV strings based on a differentiable fast fault simulation model (DFFSM). The proposed DFFSM accurately models I-V characteristics under multiple faults and provides analytical gradients with respect to fault parameters. Leveraging this property, a gradient-based fault parameters identification (GFPI) method using the Adahessian optimizer is developed to efficiently quantify partial shading, short-circuit, and series-resistance degradation. Experimental results on both simulated and measured I-V curves demonstrate that the proposed GFPI achieves high quantification accuracy across different faults, with the I-V reconstruction error below 3%, confirming the feasibility and effectiveness of the application of differentiable physical simulators for PV system fault diagnosis.
【13】How to Square Tensor Networks and Circuits Without Squaring Them
标题：如何在不平方的情况下对张量网络和电路进行平方
链接：https://arxiv.org/abs/2512.17090
摘要：平方张量网络（TN）及其作为计算图的扩展-平方电路-已被用作表达分布估计，但支持封闭形式的边缘化。然而，平方运算在计算配分函数或边缘化变量时引入了额外的复杂性，这阻碍了它们在ML中的适用性。为了解决这个问题，TN的标准型通过酉矩阵参数化，以简化边缘的计算。然而，这些标准形式不适用于电路，因为它们可以表示不直接映射到已知TN的因子分解。受启发的思想，正交的规范形式和确定性的电路，使易于处理的最大化，我们展示了如何参数化平方电路，以克服其边缘化开销。我们的参数化解锁有效的边缘化，即使在分解不同的TN，但编码为电路，其结构，否则边缘化计算困难。最后，我们在分布估计上的实验显示了我们在平方电路中提出的条件如何在没有表现力损失的情况下实现更有效的学习。
摘要：Squared tensor networks (TNs) and their extension as computational graphs--squared circuits--have been used as expressive distribution estimators, yet supporting closed-form marginalization. However, the squaring operation introduces additional complexity when computing the partition function or marginalizing variables, which hinders their applicability in ML. To solve this issue, canonical forms of TNs are parameterized via unitary matrices to simplify the computation of marginals. However, these canonical forms do not apply to circuits, as they can represent factorizations that do not directly map to a known TN. Inspired by the ideas of orthogonality in canonical forms and determinism in circuits enabling tractable maximization, we show how to parameterize squared circuits to overcome their marginalization overhead. Our parameterizations unlock efficient marginalization even in factorizations different from TNs, but encoded as circuits, whose structure would otherwise make marginalization computationally hard. Finally, our experiments on distribution estimation show how our proposed conditions in squared circuits come with no expressiveness loss, while enabling more efficient learning.
【14】SFBD-OMNI: Bridge models for lossy measurement restoration with limited clean samples
标题：SFBD-OMNI：用于有限干净样本的有损测量恢复的桥接模型
链接：https://arxiv.org/abs/2512.17051
摘要：在许多现实世界的情况下，获得完全观察到的样本是非常昂贵的，甚至是不可行的，而部分和嘈杂的意见是比较容易collect. In这项工作中，我们研究了分布恢复丰富的噪声样本，假设腐败过程是一个黑盒发电机。我们表明，这项任务可以被视为一个片面的熵最优运输问题，并通过EM类算法解决。我们进一步提供了一个测试标准，以确定是否真正的底层分布是可恢复的每个样本的信息损失下，并表明，在其他不可恢复的情况下，少量的干净的样本可以使分布在很大程度上可恢复。基于这些见解，我们引入了SFBD-OMNI，这是一个基于桥梁模型的框架，可以将损坏的样本分布映射到地面真实分布。我们的方法推广了随机前向-后向反卷积（SFBD; Lu等人，2025）处理任意测量模型超出高斯腐败。跨基准数据集和不同测量设置的实验表明，定性和定量性能都有显着改善。
摘要：In many real-world scenarios, obtaining fully observed samples is prohibitively expensive or even infeasible, while partial and noisy observations are comparatively easy to collect. In this work, we study distribution restoration with abundant noisy samples, assuming the corruption process is available as a black-box generator. We show that this task can be framed as a one-sided entropic optimal transport problem and solved via an EM-like algorithm. We further provide a test criterion to determine whether the true underlying distribution is recoverable under per-sample information loss, and show that in otherwise unrecoverable cases, a small number of clean samples can render the distribution largely recoverable. Building on these insights, we introduce SFBD-OMNI, a bridge model-based framework that maps corrupted sample distributions to the ground-truth distribution. Our method generalizes Stochastic Forward-Backward Deconvolution (SFBD; Lu et al., 2025) to handle arbitrary measurement models beyond Gaussian corruption. Experiments across benchmark datasets and diverse measurement settings demonstrate significant improvements in both qualitative and quantitative performance.
【15】Learning vertical coordinates via automatic differentiation of a dynamical core
标题：通过动态核心的自动求导学习垂直坐标
链接：https://arxiv.org/abs/2512.17877
摘要：大气模式中的地形跟踪坐标经常将其网格结构印在解上，特别是在陡峭的地形上，扭曲的坐标层会产生虚假的水平和垂直运动。标准公式，如混合或套筒坐标，通过使用由启发式尺度参数控制的分析衰减函数来减轻这些误差，这些参数通常是手动调整和先验固定的。在这项工作中，我们提出了一个框架来定义一个参数的垂直坐标系作为一个可学习的组件内的可微动态核心。我们开发了一个端到端的可微数值求解器的二维非静力欧拉方程的Arakawa C-网格，并引入了一个神经垂直增强（NEURAL垂直增强）地形跟踪坐标的基础上，保证单调性的积分变换神经网络。我们的方法的一个关键特征是使用自动微分计算精确的几何度量项，从而消除与有限差分坐标导数相关的截断误差。通过将模拟误差通过时间积分耦合到参数化，我们的公式找到了针对底层物理和数值优化的网格结构。使用几个标准的测试，我们证明，这些学习的坐标减少了1.4至2的非线性统计基准的因素的均方误差，并消除虚假的垂直速度条纹陡峭的地形。
摘要：Terrain-following coordinates in atmospheric models often imprint their grid structure onto the solution, particularly over steep topography, where distorted coordinate layers can generate spurious horizontal and vertical motion. Standard formulations, such as hybrid or SLEVE coordinates, mitigate these errors by using analytic decay functions controlled by heuristic scale parameters that are typically tuned by hand and fixed a priori. In this work, we propose a framework to define a parametric vertical coordinate system as a learnable component within a differentiable dynamical core. We develop an end-to-end differentiable numerical solver for the two-dimensional non-hydrostatic Euler equations on an Arakawa C-grid, and introduce a NEUral Vertical Enhancement (NEUVE) terrain-following coordinate based on an integral transformed neural network that guarantees monotonicity. A key feature of our approach is the use of automatic differentiation to compute exact geometric metric terms, thereby eliminating truncation errors associated with finite-difference coordinate derivatives. By coupling simulation errors through the time integration to the parameterization, our formulation finds a grid structure optimized for both the underlying physics and numerics. Using several standard tests, we demonstrate that these learned coordinates reduce the mean squared error by a factor of 1.4 to 2 in non-linear statistical benchmarks, and eliminate spurious vertical velocity striations over steep topography.
【16】Revisiting the Broken Symmetry Phase of Solid Hydrogen: A Neural Network Variational Monte Carlo Study
标题：重新审视固体氢的对称性破碎阶段：神经网络变分蒙特卡罗研究
链接：https://arxiv.org/abs/2512.17703
摘要：高压固态氢的晶体结构仍然是一个基本的悬而未决的问题。虽然研究前沿主要转向400 GPa以上的超高压阶段，我们表明，即使是在130 GPa左右观察到的对称性破缺阶段，由于其复杂的电子和核自由度的耦合，需要重新审视。在这里，我们开发了一个基于深度神经网络波函数的第一原理量子蒙特卡罗框架，该框架在恒压系综中以量子力学的方式处理电子和原子核。我们的计算揭示了一个未报道的基态结构候选人的对称性破缺相与$Cmcm$空间群对称性，我们测试其稳定性高达96个原子。预测的结构定量匹配的实验状态方程和X射线衍射图案。此外，我们的群论分析表明，$Cmcm$结构与现有的拉曼和红外光谱数据是兼容的。至关重要的是，静态密度泛函理论计算揭示了$Cmcm$结构作为一个动态不稳定的鞍点上的Born-Oppenheimer势能面，表明一个完整的量子多体处理的问题是必要的。这些结果揭示了新的光的高压氢的相图，并呼吁进一步的实验验证。
摘要 ：The crystal structure of high-pressure solid hydrogen remains a fundamental open problem. Although the research frontier has mostly shifted toward ultra-high pressure phases above 400 GPa, we show that even the broken symmetry phase observed around 130~GPa requires revisiting due to its intricate coupling of electronic and nuclear degrees of freedom. Here, we develop a first principle quantum Monte Carlo framework based on a deep neural network wave function that treats both electrons and nuclei quantum mechanically within the constant pressure ensemble. Our calculations reveal an unreported ground-state structure candidate for the broken symmetry phase with $Cmcm$ space group symmetry, and we test its stability up to 96 atoms. The predicted structure quantitatively matches the experimental equation of state and X-ray diffraction patterns. Furthermore, our group-theoretical analysis shows that the $Cmcm$ structure is compatible with existing Raman and infrared spectroscopic data. Crucially, static density functional theory calculation reveals the $Cmcm$ structure as a dynamically unstable saddle point on the Born-Oppenheimer potential energy surface, demonstrating that a full quantum many-body treatment of the problem is necessary. These results shed new light on the phase diagram of high-pressure hydrogen and call for further experimental verifications.
【17】SkinGenBench: Generative Model and Preprocessing Effects for Synthetic Dermoscopic Augmentation in Melanoma Diagnosis
标题：SkinGenBench：黑色素瘤诊断中合成皮肤镜增强的生成模型和预处理效果
链接：https://arxiv.org/abs/2512.17585
摘要：这项工作介绍了SkinGenBench，这是一个系统的生物医学成像基准，研究了预处理复杂性如何与合成皮肤镜图像增强和下游黑色素瘤诊断的生成模型选择相互作用。使用来自HAM 10000和MILK 10 K的14，116张皮肤镜图像的策划数据集，我们评估了两种代表性的生成范例：StyleGAN 2-ADA和基本几何增强和高级伪影去除管道下的去噪扩散概率模型（DDPM）。使用已建立的感知和分布指标（FID，KID，IS），特征空间分析及其对五个下游分类器诊断性能的影响来评估合成黑色素瘤图像。实验结果表明，生成架构的选择有更强的影响，图像保真度和诊断效用比预处理复杂性。StyleGAN 2-ADA始终生成与真实数据分布更接近的合成图像，实现了最低的FID（~65.5）和KID（~0.05），而扩散模型生成了更高的方差样本，代价是降低了感知保真度和类锚定。先进的伪影去除只产生了边际改进生成指标，并提供了有限的下游诊断收益，这表明可能抑制临床相关的纹理线索。相比之下，合成数据增强显著改善了黑色素瘤检测，黑色素瘤F1评分的绝对增益为8-15%，ViT-B/16实现了F1~0.88和ROC-AUC~0.98，与非增强基线相比改善了约14%。我们的代码可在https://github.com/adarsh-crafts/SkinGenBench上找到
摘要：This work introduces SkinGenBench, a systematic biomedical imaging benchmark that investigates how preprocessing complexity interacts with generative model choice for synthetic dermoscopic image augmentation and downstream melanoma diagnosis. Using a curated dataset of 14,116 dermoscopic images from HAM10000 and MILK10K across five lesion classes, we evaluate the two representative generative paradigms: StyleGAN2-ADA and Denoising Diffusion Probabilistic Models (DDPMs) under basic geometric augmentation and advanced artifact removal pipelines. Synthetic melanoma images are assessed using established perceptual and distributional metrics (FID, KID, IS), feature space analysis, and their impact on diagnostic performance across five downstream classifiers. Experimental results demonstrate that generative architecture choice has a stronger influence on both image fidelity and diagnostic utility than preprocessing complexity. StyleGAN2-ADA consistently produced synthetic images more closely aligned with real data distributions, achieving the lowest FID (~65.5) and KID (~0.05), while diffusion models generated higher variance samples at the cost of reduces perceptual fidelity and class anchoring. Advanced artifact removal yielded only marginal improvements in generative metrics and provided limited downstream diagnostic gains, suggesting possible suppression of clinically relevant texture cues. In contrast, synthetic data augmentation substantially improved melanoma detection with 8-15% absolute gains in melanoma F1-score, and ViT-B/16 achieving F1~0.88 and ROC-AUC~0.98, representing an improvement of approximately 14% over non-augmented baselines. Our code can be found at https://github.com/adarsh-crafts/SkinGenBench
【18】Machine Learning Assisted Parameter Tuning on Wavelet Transform Amorphous Radial Distribution Function
标题：基于机器学习的小波变换非晶径向分布函数参数整定
链接：https://arxiv.org/abs/2512.17245
摘要：了解原子结构至关重要，但由于其不规则和非周期性，非晶材料仍然具有挑战性。小波变换径向分布函数（WT-RDF）为分析非晶结构提供了一个基于物理的框架，可靠地预测了二元Ge 0.25 Se 0.75和三元Ag x（Ge 0.25 Se 0.75）100-x（x= 5，10，15，20，25）系统的第一和第二RDF峰以及整体曲线趋势。尽管有这些优势，WT-RDF显示振幅精度的限制，这影响了定量分析，如协调数。本研究通过使用机器学习方法优化WT-RDF参数来解决这个问题，产生增强的WT-RDF+框架。WT-RDF+提高了峰值预测的精度，甚至在只对25%的二进制数据集进行训练时，性能也优于基准ML模型，包括RBF和LSTM。这些结果表明，WT-RDF+是一个强大的和可靠的非晶材料，特别是Ge-Se系统的结构表征模型，并支持下一代电子器件和组件的相变薄膜的有效设计和开发。
摘要：Understanding atomic structures is crucial, yet amorphous materials remain challenging due to their irregular and non-periodic nature. The wavelet-transform radial distribution function (WT-RDF) offers a physics-based framework for analyzing amorphous structures, reliably predicting the first and second RDF peaks and overall curve trends in both binary Ge 0.25 Se 0.75 and ternary Ag x(Ge 0.25 Se 0.75)100-x (x=5,10,15,20,25) systems. Despite these strengths, WT-RDF shows limitations in amplitude accuracy, which affects quantitative analyses such as coordination numbers. This study addresses the issue by optimizing WT-RDF parameters using a machine learning approach, producing the enhanced WT-RDF+ framework. WT-RDF+ improves the precision of peak predictions and outperforms benchmark ML models, including RBF and LSTM, even when trained on only 25 percent of the binary dataset. These results demonstrate that WT-RDF+ is a robust and reliable model for structural characterization of amorphous materials, particularly Ge-Se systems, and support the efficient design and development of phase-change thin films for next-generation electronic devices and components.
【19】Application of machine learning to predict food processing level using Open Food Facts
标题：应用机器学习使用Open Food Facts预测食品加工水平
链接：https://arxiv.org/abs/2512.17169
备注：27 Pages (22 Pages of Main Manuscript + Supplementary Material), 7 Figures, 1 Table
摘要：由于营养质量差，超加工食品越来越多地与肥胖，心血管疾病，2型糖尿病和精神健康障碍等健康问题联系在一起。这项规模空前的研究使用机器学习来对食品加工水平（NOVA）进行分类，该数据集基于超过90万种产品的Open Food Facts数据集。包括LightGBM，Random Forest和CatBoost在内的模型都是在营养浓度数据上训练的。LightGBM表现最好，在不同的营养成分面板上达到80-85%的准确率，并有效地将其与超加工食品进行最小化区分。探索性分析显示，较高的NOVA类和较低的营养分数之间存在很强的关联，表明营养质量较差。NOVA 3和4中的产品也具有较高的碳足迹和较低的生态评分，表明对环境的影响更大。过敏原分析发现，面筋和牛奶在超加工食品中很常见，对敏感人群构成风险。蛋糕和零食等类别在较高的NOVA类别中占主导地位，这些类别也有更多的添加剂，突出了成分修改的作用。这项研究利用了NOVA标记产品的最大数据集，强调了食品加工的健康，环境和过敏影响，并展示了机器学习在可扩展分类中的价值。一个用户友好的网络工具可用于使用营养数据进行NOVA预测：https://cosylab.iiitd.edu.in/foodlabel/。
摘要：Ultra-processed foods are increasingly linked to health issues like obesity, cardiovascular disease, type 2 diabetes, and mental health disorders due to poor nutritional quality. This first-of-its-kind study at such a scale uses machine learning to classify food processing levels (NOVA) based on the Open Food Facts dataset of over 900,000 products. Models including LightGBM, Random Forest, and CatBoost were trained on nutrient concentration data. LightGBM performed best, achieving 80-85% accuracy across different nutrient panels and effectively distinguishing minimally from ultra-processed foods. Exploratory analysis revealed strong associations between higher NOVA classes and lower Nutri-Scores, indicating poorer nutritional quality. Products in NOVA 3 and 4 also had higher carbon footprints and lower Eco-Scores, suggesting greater environmental impact. Allergen analysis identified gluten and milk as common in ultra-processed items, posing risks to sensitive individuals. Categories like Cakes and Snacks were dominant in higher NOVA classes, which also had more additives, highlighting the role of ingredient modification. This study, leveraging the largest dataset of NOVA-labeled products, emphasizes the health, environmental, and allergenic implications of food processing and showcases machine learning's value in scalable classification. A user-friendly web tool is available for NOVA prediction using nutrient data: https://cosylab.iiitd.edu.in/foodlabel/.
其他(26篇)
【1】Weighted Stochastic Differential Equation to Implement Wasserstein-Fisher-Rao Gradient Flow
标题：加权随机方程实现Wasserstein-Fisher-Rao梯度流
链接：https://arxiv.org/abs/2512.17878
备注：26 pages, 1 figure
摘要：基于分数的扩散模型目前构成了连续生成建模的最新技术。这些方法通常通过过阻尼或欠阻尼Ornstein-Uhlenbeck型随机微分方程来制定，其中采样由确定性漂移和布朗扩散的组合来驱动，从而在环境空间中产生连续的粒子轨迹。虽然这样的动态享受指数收敛保证强对数凹目标分布，它是众所周知的，它们的混合率指数恶化的存在下，非凸或多峰景观，如双阱势。由于许多实际的生成建模任务涉及高度非对数凹目标分布，相当多的最近的努力一直致力于开发采样方案，提高探索超越经典的扩散动力学。 一个有前途的工作线利用信息几何工具，以增加基于扩散的采样器与控制质量重加权机制。这种观点自然导致Wasserstein-Fisher-Rao（WFR）几何形状，它耦合在样品空间中的运输与垂直（反应）的概率测度空间上的动态。在这项工作中，我们制定这样的重新加权机制，通过引入明确的校正项，并显示它们如何可以通过加权随机微分方程使用费曼-卡茨表示。我们的研究提供了一个初步的，但严格的调查基于WFR的采样动力学，并旨在澄清其几何和运营商的理论结构作为未来的理论和算法的发展的基础。
摘要：Score-based diffusion models currently constitute the state of the art in continuous generative modeling. These methods are typically formulated via overdamped or underdamped Ornstein--Uhlenbeck-type stochastic differential equations, in which sampling is driven by a combination of deterministic drift and Brownian diffusion, resulting in continuous particle trajectories in the ambient space. While such dynamics enjoy exponential convergence guarantees for strongly log-concave target distributions, it is well known that their mixing rates deteriorate exponentially in the presence of nonconvex or multimodal landscapes, such as double-well potentials. Since many practical generative modeling tasks involve highly non-log-concave target distributions, considerable recent effort has been devoted to developing sampling schemes that improve exploration beyond classical diffusion dynamics. A promising line of work leverages tools from information geometry to augment diffusion-based samplers with controlled mass reweighting mechanisms. This perspective leads naturally to Wasserstein--Fisher--Rao (WFR) geometries, which couple transport in the sample space with vertical (reaction) dynamics on the space of probability measures. In this work, we formulate such reweighting mechanisms through the introduction of explicit correction terms and show how they can be implemented via weighted stochastic differential equations using the Feynman--Kac representation. Our study provides a preliminary but rigorous investigation of WFR-based sampling dynamics, and aims to clarify their geometric and operator-theoretic structure as a foundation for future theoretical and algorithmic developments.
【2】Visually Prompted Benchmarks Are Surprisingly Fragile
标题：视觉上的基准出奇地脆弱
链接：https://arxiv.org/abs/2512.17875
摘要：评估VLM的一个关键挑战是测试模型独立于其文本先验分析视觉内容的能力。最近的基准测试，如BLINK探测视觉感知通过视觉提示，其中关于视觉内容的问题与问题所指的坐标配对，坐标在图像本身中明确标记。虽然这些基准测试是VLM评估的重要组成部分，但我们发现现有模型对视觉提示的看似无关的细节非常脆弱：简单地将视觉标记从红色更改为蓝色可以完全改变排行榜上模型之间的排名。通过在两个视觉提示任务上评估九个常用的开源和闭源VLM，我们展示了基准设置中的细节，包括视觉标记设计和数据集大小，对模型性能和排行榜排名有显着影响。这些效果甚至可以被利用来提升较弱的模型，使其高于较强的模型;例如，稍微增加视觉标记的大小会导致开源InternVL 3 -8B的排名与Gemini 2.5 Pro等更大的专有模型并列或更好。我们进一步表明，在基准测试中经常被忽略的低级推理选择，例如API调用中的JPEG压缩级别，也可能导致模型阵容的变化。这些细节对视觉提示基准的影响比传统的语义VLM评估大得多。为了减轻这种不稳定性，我们整理现有的数据集来创建VPBench，这是一个更大的视觉提示基准，具有16个视觉标记变体。VPBench和其他分析工具在https://lisadunlap.github.io/vpbench/上发布。
摘要：A key challenge in evaluating VLMs is testing models' ability to analyze visual content independently from their textual priors. Recent benchmarks such as BLINK probe visual perception through visual prompting, where questions about visual content are paired with coordinates to which the question refers, with the coordinates explicitly marked in the image itself. While these benchmarks are an important part of VLM evaluation, we find that existing models are surprisingly fragile to seemingly irrelevant details of visual prompting: simply changing a visual marker from red to blue can completely change rankings among models on a leaderboard. By evaluating nine commonly-used open- and closed-source VLMs on two visually prompted tasks, we demonstrate how details in benchmark setup, including visual marker design and dataset size, have a significant influence on model performance and leaderboard rankings. These effects can even be exploited to lift weaker models above stronger ones; for instance, slightly increasing the size of the visual marker results in open-source InternVL3-8B ranking alongside or better than much larger proprietary models like Gemini 2.5 Pro. We further show that low-level inference choices that are often ignored in benchmarking, such as JPEG compression levels in API calls, can also cause model lineup changes. These details have substantially larger impacts on visually prompted benchmarks than on conventional semantic VLM evaluations. To mitigate this instability, we curate existing datasets to create VPBench, a larger visually prompted benchmark with 16 visual marker variants. VPBench and additional analysis tools are released at https://lisadunlap.github.io/vpbench/.
【3】You Only Train Once: Differentiable Subset Selection for Omics Data
标题：您只能训练一次：组学数据的差异子集选择
链接：https://arxiv.org/abs/2512.17678
摘要：从单细胞转录组数据中选择紧凑且信息丰富的基因子集对于生物标志物发现、提高可解释性和成本效益分析至关重要。然而，大多数现有的特征选择方法要么作为多阶段管道操作，要么依赖于事后特征属性，使得选择和预测弱耦合。在这项工作中，我们提出了YOTO（你只需要训练一次），这是一个端到端的框架，它可以联合识别离散的基因子集，并在一个可区分的架构中进行预测。在我们的模型中，预测任务直接指导选择哪些基因，而学习的子集反过来又塑造了预测表示。这种闭环反馈使模型能够迭代地优化它在训练过程中选择的内容和预测方式。与现有的方法不同，YOTO强制执行稀疏性，因此只有选定的基因有助于推理，无需训练额外的下游分类器。通过多任务学习设计，该模型学习相关目标之间的共享表示，允许部分标记的数据集相互通知，并发现跨任务泛化的基因子集，而无需额外的训练步骤。我们在两个具有代表性的单细胞RNA-seq数据集上评估了YOTO，结果表明它的性能始终优于最先进的基线。这些结果表明，稀疏、端到端、多任务基因子集选择提高了预测性能，并产生了紧凑和有意义的基因子集，推进了生物标志物发现和单细胞分析。
摘要 ：Selecting compact and informative gene subsets from single-cell transcriptomic data is essential for biomarker discovery, improving interpretability, and cost-effective profiling. However, most existing feature selection approaches either operate as multi-stage pipelines or rely on post hoc feature attribution, making selection and prediction weakly coupled. In this work, we present YOTO (you only train once), an end-to-end framework that jointly identifies discrete gene subsets and performs prediction within a single differentiable architecture. In our model, the prediction task directly guides which genes are selected, while the learned subsets, in turn, shape the predictive representation. This closed feedback loop enables the model to iteratively refine both what it selects and how it predicts during training. Unlike existing approaches, YOTO enforces sparsity so that only the selected genes contribute to inference, eliminating the need to train additional downstream classifiers. Through a multi-task learning design, the model learns shared representations across related objectives, allowing partially labeled datasets to inform one another, and discovering gene subsets that generalize across tasks without additional training steps. We evaluate YOTO on two representative single-cell RNA-seq datasets, showing that it consistently outperforms state-of-the-art baselines. These results demonstrate that sparse, end-to-end, multi-task gene subset selection improves predictive performance and yields compact and meaningful gene subsets, advancing biomarker discovery and single-cell analysis.
【4】Polyharmonic Cascade
标题：多调和级联
链接：https://arxiv.org/abs/2512.17671
备注：Part 3 of 4 in the "Polyharmonic Cascade" cycle. Proposes a non-SGD training method based on global linear solvers. Previous papers: arXiv:2512.12731, arXiv.2512.16718. Source code is available at: https://github.com/xolod7/polyharmonic-cascade
摘要：本文提出了一种深度机器学习架构，即“多谐级联”--一系列多谐样条包，其中每一层都严格源自随机函数理论和无差异原理。这使得可以近似任意复杂度的非线性函数，同时保持全局光滑性和概率解释。对于多谐级联，提出了一种替代梯度下降的训练方法：而不是直接优化系数，而是在每个批次上相对于固定节点“星座”处的函数值求解单个全局线性系统。这产生了所有层的同步更新，保留了单个层的概率解释和与原始模型的理论一致性，并且扩展良好：所有计算都简化为在GPU上有效执行的2D矩阵运算。MNIST上的快速学习没有过拟合证明。
摘要：This paper presents a deep machine learning architecture, the "polyharmonic cascade" -- a sequence of packages of polyharmonic splines, where each layer is rigorously derived from the theory of random functions and the principles of indifference. This makes it possible to approximate nonlinear functions of arbitrary complexity while preserving global smoothness and a probabilistic interpretation. For the polyharmonic cascade, a training method alternative to gradient descent is proposed: instead of directly optimizing the coefficients, one solves a single global linear system on each batch with respect to the function values at fixed "constellations" of nodes. This yields synchronized updates of all layers, preserves the probabilistic interpretation of individual layers and theoretical consistency with the original model, and scales well: all computations reduce to 2D matrix operations efficiently executed on a GPU. Fast learning without overfitting on MNIST is demonstrated.
【5】More Consistent Accuracy PINN via Alternating Easy-Hard Training
标题：通过交替易难训练提高准确性PINN
链接：https://arxiv.org/abs/2512.17607
摘要：物理信息神经网络（PINN）最近已经成为解决偏微分方程（PDE）的一个突出范例，但其训练策略仍然未得到充分研究。虽然受有限元方法启发的硬优先级方法被广泛采用，但最近的研究表明，简单的优先级排序也可以有效。然而，我们发现，这两种方法表现出显着的权衡和不一致的性能在PDE类型。为了解决这个问题，我们开发了一种混合策略，通过交替训练算法结合了硬优先级和简单优先级的优势。在具有陡梯度、非线性和高维的偏微分方程上，该方法实现了一致的高精度，相对L2误差大多在O（10^-5）到O（10^-6）的范围内，显著优于基线方法。此外，它提供了更大的可靠性，在不同的问题，而比较的方法往往遭受变量的精度取决于PDE。这项工作为设计混合训练策略以提高PINN的性能和鲁棒性提供了新的见解。
摘要：Physics-informed neural networks (PINNs) have recently emerged as a prominent paradigm for solving partial differential equations (PDEs), yet their training strategies remain underexplored. While hard prioritization methods inspired by finite element methods are widely adopted, recent research suggests that easy prioritization can also be effective. Nevertheless, we find that both approaches exhibit notable trade-offs and inconsistent performance across PDE types. To address this issue, we develop a hybrid strategy that combines the strengths of hard and easy prioritization through an alternating training algorithm. On PDEs with steep gradients, nonlinearity, and high dimensionality, the proposed method achieves consistently high accuracy, with relative L2 errors mostly in the range of O(10^-5) to O(10^-6), significantly surpassing baseline methods. Moreover, it offers greater reliability across diverse problems, whereas compared approaches often suffer from variable accuracy depending on the PDE. This work provides new insights into designing hybrid training strategies to enhance the performance and robustness of PINNs.
【6】Bayesian Optimisation: Which Constraints Matter?
标题：Bayesian优化：哪些约束重要？
链接：https://arxiv.org/abs/2512.17569
摘要：贝叶斯优化已被证明是一个强大的工具，昂贵的全球黑箱优化问题。在本文中，我们提出了新的贝叶斯优化变种流行的知识梯度采集功能的问题与\n {解耦}黑盒约束，其中的目标和约束函数的子集可以独立评估。特别是，我们的方法的目的是考虑到，往往只有少数的约束可能是约束力的最佳，因此，我们应该只评估相关的约束时，试图优化功能。我们经验基准这些方法对现有的方法，并证明其优越性的国家的最先进的。
摘要：Bayesian optimisation has proven to be a powerful tool for expensive global black-box optimisation problems. In this paper, we propose new Bayesian optimisation variants of the popular Knowledge Gradient acquisition functions for problems with \emph{decoupled} black-box constraints, in which subsets of the objective and constraint functions may be evaluated independently. In particular, our methods aim to take into account that often only a handful of the constraints may be binding at the optimum, and hence we should evaluate only relevant constraints when trying to optimise a function. We empirically benchmark these methods against existing methods and demonstrate their superiority over the state-of-the-art.
【7】Translating the Rashomon Effect to Sequential Decision-Making Tasks
标题：将罗生门效应转化为顺序决策任务
链接：https://arxiv.org/abs/2512.17470
摘要：罗生门效应描述了一种现象，即在相同数据上训练的多个模型产生相同的预测，但它们内部依赖的特征不同。这种效应在分类任务中得到了广泛的研究，但在顺序决策中却没有，在顺序决策中，智能体通过在环境中采取行动来学习策略以实现目标。在本文中，我们将罗生门效应转化为序列决策。我们将其定义为表现出相同行为的多个策略，访问相同的状态并选择相同的操作，但其内部结构不同，例如特征属性。顺序决策中的完全相同行为不同于分类。在分类中，预测可以直接与地面实况标签进行比较。在具有随机转移的顺序决策中，由于随机性，同一策略可能在任何单个轨迹上成功或失败。我们解决这个问题，使用正式的验证方法，构建和比较环境中的每个政策的完整的概率行为。我们的实验表明，罗生门效应存在于序列决策。我们进一步表明，从罗生门集构造的合奏表现出更大的鲁棒性比个人的政策分布变化。此外，从罗生门集导出的许可策略减少了验证的计算需求，同时保持了最佳性能。
摘要 ：The Rashomon effect describes the phenomenon where multiple models trained on the same data produce identical predictions while differing in which features they rely on internally. This effect has been studied extensively in classification tasks, but not in sequential decision-making, where an agent learns a policy to achieve an objective by taking actions in an environment. In this paper, we translate the Rashomon effect to sequential decision-making. We define it as multiple policies that exhibit identical behavior, visiting the same states and selecting the same actions, while differing in their internal structure, such as feature attributions. Verifying identical behavior in sequential decision-making differs from classification. In classification, predictions can be directly compared to ground-truth labels. In sequential decision-making with stochastic transitions, the same policy may succeed or fail on any single trajectory due to randomness. We address this using formal verification methods that construct and compare the complete probabilistic behavior of each policy in the environment. Our experiments demonstrate that the Rashomon effect exists in sequential decision-making. We further show that ensembles constructed from the Rashomon set exhibit greater robustness to distribution shifts than individual policies. Additionally, permissive policies derived from the Rashomon set reduce computational requirements for verification while maintaining optimal performance.
【8】Behavioural Effects of Agentic Messaging: A Case Study on a Financial Service Application
标题：广告信息的行为影响：金融服务应用程序的案例研究
链接：https://arxiv.org/abs/2512.17462
备注：To appear in the 48th European Conference on Information Retrieval (ECIR '26) Industry Track
摘要：营销和产品个性化为跨多个业务领域的信息检索方法的应用提供了突出和可见的用例。最近，对这些问题的代理方法越来越受欢迎。这项工作评估了在2025年国家税务申报期内，代理个性化对金融服务应用程序客户通信系统的行为和保留效果。通过为期两个月的随机对照试验，我们将代理消息传递方法与基于规则的活动系统进行了比较，重点关注两个主要结果：退订行为和转换时间。实证结果表明，代理人主导的消息减少退订事件的21%（$\pm 0.01 $）相对于BAU和增加早期提交行为在国家截止日期前几周。这些发现表明，自适应的用户级决策系统如何调节参与强度，同时改善长期保留指标。
摘要：Marketing and product personalisation provide a prominent and visible use-case for the application of Information Retrieval methods across several business domains. Recently, agentic approaches to these problems have been gaining traction. This work evaluates the behavioural and retention effects of agentic personalisation on a financial service application's customer communication system during a 2025 national tax filing period. Through a two month-long randomised controlled trial, we compare an agentic messaging approach against a business-as-usual (BAU) rule-based campaign system, focusing on two primary outcomes: unsubscribe behaviour and conversion timing. Empirical results show that agent-led messaging reduced unsubscribe events by 21\% ($\pm 0.01$) relative to BAU and increased early filing behaviour in the weeks preceding the national deadline. These findings demonstrate how adaptive, user-level decision-making systems can modulate engagement intensity whilst improving long-term retention indicators.
【9】MULTIAQUA: A multimodal maritime dataset and robust training strategies for multimodal semantic segmentation
标题：MULTIAQUA：多模式海事数据集和用于多模式语义分割的稳健训练策略
链接：https://arxiv.org/abs/2512.17450
摘要：无人水面航行器在运行过程中可能会遇到许多不同的视觉环境，其中一些可能非常难以解释。虽然大多数情况下只能使用彩色相机图像来解决，但某些天气和照明条件需要额外的信息。为了扩展现有的海洋数据，我们提出了一种新的多模态海洋数据集MULTIAQUA（多模态水生数据集）。我们的数据集包含由不同模态的传感器捕获的同步，校准和注释数据，例如RGB，热，IR，LIDAR等数据集旨在开发监督方法，可以从这些模态中提取有用的信息，以便提供高质量的场景解释，而不管潜在的能见度差的条件。为了说明所提出的数据集的好处，我们评估了几个多模态方法在我们困难的夜间测试集。我们提出了训练方法，使多模态方法能够以更强大的方式进行训练，从而使它们即使在接近完全黑暗的情况下也能保持可靠的性能。我们的方法允许只使用白天的图像来训练一个强大的深度神经网络，从而大大简化了数据采集、注释和训练过程。
摘要：Unmanned surface vehicles can encounter a number of varied visual circumstances during operation, some of which can be very difficult to interpret. While most cases can be solved only using color camera images, some weather and lighting conditions require additional information. To expand the available maritime data, we present a novel multimodal maritime dataset MULTIAQUA (Multimodal Aquatic Dataset). Our dataset contains synchronized, calibrated and annotated data captured by sensors of different modalities, such as RGB, thermal, IR, LIDAR, etc. The dataset is aimed at developing supervised methods that can extract useful information from these modalities in order to provide a high quality of scene interpretation regardless of potentially poor visibility conditions. To illustrate the benefits of the proposed dataset, we evaluate several multimodal methods on our difficult nighttime test set. We present training approaches that enable multimodal methods to be trained in a more robust way, thus enabling them to retain reliable performance even in near-complete darkness. Our approach allows for training a robust deep neural network only using daytime images, thus significantly simplifying data acquisition, annotation, and the training process.
【10】Timely Information Updating for Mobile Devices Without and With ML Advice
标题：在没有和有ML建议的情况下为移动设备及时更新信息
链接：https://arxiv.org/abs/2512.17381
备注：23 pages, journal version of arXiv:1901.03137, submitted for possible journal publication
摘要：本文研究了一种信息更新系统，其中移动终端监视物理过程并向接入点（AP）发送状态更新。在AP处维护的信息的及时性与在设备处发生的更新成本之间出现基本的权衡。为了解决这个权衡，我们提出了一个在线算法，确定何时传输更新只使用可用的观察。所提出的算法渐近达到最佳的竞争比对对手，可以同时操纵多个来源的不确定性，包括操作持续时间，信息陈旧，更新成本，和更新机会的可用性。此外，通过将未知可靠性的机器学习（ML）建议纳入设计中，我们开发了一种ML增强算法，该算法渐进地达到最佳的一致性-鲁棒性权衡，即使对手可以额外损坏ML建议。最优竞争比与更新成本的范围呈线性关系，但不受其他不确定性的影响。此外，最优竞争在线算法对ML建议表现出类似阈值的响应：它完全信任或完全忽略ML建议，因为部分信任建议无法在不严重降低鲁棒性的情况下提高一致性。随机环境中的大量模拟进一步验证了对抗环境中的理论发现。
摘要：This paper investigates an information update system in which a mobile device monitors a physical process and sends status updates to an access point (AP). A fundamental trade-off arises between the timeliness of the information maintained at the AP and the update cost incurred at the device. To address this trade-off, we propose an online algorithm that determines when to transmit updates using only available observations. The proposed algorithm asymptotically achieves the optimal competitive ratio against an adversary that can simultaneously manipulate multiple sources of uncertainty, including the operation duration, the information staleness, the update cost, and the availability of update opportunities. Furthermore, by incorporating machine learning (ML) advice of unknown reliability into the design, we develop an ML-augmented algorithm that asymptotically attains the optimal consistency-robustness trade-off, even when the adversary can additionally corrupt the ML advice. The optimal competitive ratio scales linearly with the range of update costs, but is unaffected by other uncertainties. Moreover, an optimal competitive online algorithm exhibits a threshold-like response to the ML advice: it either fully trusts or completely ignores the ML advice, as partially trusting the advice cannot improve the consistency without severely degrading the robustness. Extensive simulations in stochastic settings further validate the theoretical findings in the adversarial environment.
【11】Explanation Beyond Intuition: A Testable Criterion for Inherent Explainability
标题：超越直觉的解释：内在可解释性的可测试标准
链接：https://arxiv.org/abs/2512.17316
摘要 ：内在可解释性是可解释人工智能（XAI）的黄金标准。然而，没有一致的定义或测试来证明内在的可解释性。迄今为止的工作要么通过度量来描述可解释性，要么诉诸直觉-“当我们看到它时，我们知道它”。我们提出了一个全球适用的标准，内在的可解释性。该标准使用图论来表示和分解模型以进行结构局部解释，并将其重组为全局解释。我们将结构-局部解释形成为注释，这是一种可验证的假设-证据结构，允许使用一系列解释方法。这个标准与现有的内在可解释性直觉相匹配，并为为什么大型回归模型可能无法解释而稀疏神经网络可以解释提供了理由。我们区分可解释的--一个允许解释的模型--和\textit{explained} --一个有验证的解释的模型。最后，我们提供了一个完整的解释预测-Cox比例风险模型的心血管疾病的风险，这是在积极的临床应用在新西兰。因此，预测是内在可解释的。这项工作提供了一个结构，以正式确定其他工作的可解释性，并允许监管机构灵活但严格的测试，可用于合规框架。
摘要：Inherent explainability is the gold standard in Explainable Artificial Intelligence (XAI). However, there is not a consistent definition or test to demonstrate inherent explainability. Work to date either characterises explainability through metrics, or appeals to intuition - "we know it when we see it". We propose a globally applicable criterion for inherent explainability. The criterion uses graph theory for representing and decomposing models for structure-local explanation, and recomposing them into global explanations. We form the structure-local explanations as annotations, a verifiable hypothesis-evidence structure that allows for a range of explanatory methods to be used. This criterion matches existing intuitions on inherent explainability, and provides justifications why a large regression model may not be explainable but a sparse neural network could be. We differentiate explainable -- a model that allows for explanation -- and \textit{explained} -- one that has a verified explanation. Finally, we provide a full explanation of PREDICT -- a Cox proportional hazards model of cardiovascular disease risk, which is in active clinical use in New Zealand. It follows that PREDICT is inherently explainable. This work provides structure to formalise other work on explainability, and allows regulators a flexible but rigorous test that can be used in compliance frameworks.
【12】MINPO: Memory-Informed Neural Pseudo-Operator to Resolve Nonlocal Spatiotemporal Dynamics
标题：MINPO：用于解决非局部时空动态的记忆神经伪运算符
链接：https://arxiv.org/abs/2512.17273
摘要：许多物理系统表现出非局部时空行为的积分微分方程（IDE）描述。求解IDE的经典方法需要反复计算卷积积分，其成本随着内核复杂度和维数的增加而迅速增加。现有的神经求解器可以加速这些计算的选定实例，但它们不能推广到各种非局部结构。在这项工作中，我们介绍了记忆知情的神经伪算子（MINPO），一个统一的框架，用于建模非局部动态所产生的远程空间相互作用和/或长期的时间记忆。MINPO采用Kolmogorov-Arnold网络（KANs）或多层感知器网络（MLP）作为编码器，通过神经表示直接学习非局部算子及其逆算子，然后显式重构未知解域。学习由一个轻量级的非局部一致性损失项来保护，以加强学习算子和重构解之间的一致性。MINPO制定允许自然地捕获和有效地解决非本地时空依赖性所管辖的广泛的IDE及其子集，包括分数PDE。我们评估了MINPO与经典技术和基于MLP的最先进的基于神经的策略（如A-PINN和fPINN）以及其新开发的KAN变体A-PIKAN和fPIKAN的有效性，旨在促进公平比较。我们的研究提供了令人信服的证据MINPO的准确性，并证明其鲁棒性处理（i）不同的内核类型，（ii）不同的内核维数，（iii）大量的计算需求所产生的内核积分的重复评估。MINPO，因此，概括超出了特定问题的配方，提供了一个统一的框架，由非本地运营商管理的系统。
摘要：Many physical systems exhibit nonlocal spatiotemporal behaviors described by integro-differential equations (IDEs). Classical methods for solving IDEs require repeatedly evaluating convolution integrals, whose cost increases quickly with kernel complexity and dimensionality. Existing neural solvers can accelerate selected instances of these computations, yet they do not generalize across diverse nonlocal structures. In this work, we introduce the Memory-Informed Neural Pseudo-Operator (MINPO), a unified framework for modeling nonlocal dynamics arising from long-range spatial interactions and/or long-term temporal memory. MINPO, employing either Kolmogorov-Arnold Networks (KANs) or multilayer perceptron networks (MLPs) as encoders, learns the nonlocal operator and its inverse directly through neural representations, and then explicitly reconstruct the unknown solution fields. The learning is guarded by a lightweight nonlocal consistency loss term to enforce coherence between the learned operator and reconstructed solution. The MINPO formulation allows to naturally capture and efficiently resolve nonlocal spatiotemporal dependencies governed by a wide spectrum of IDEs and their subsets, including fractional PDEs. We evaluate the efficacy of MINPO in comparison with classical techniques and state-of-the-art neural-based strategies based on MLPs, such as A-PINN and fPINN, along with their newly-developed KAN variants, A-PIKAN and fPIKAN, designed to facilitate a fair comparison. Our study offers compelling evidence of the accuracy of MINPO and demonstrates its robustness in handling (i) diverse kernel types, (ii) different kernel dimensionalities, and (iii) the substantial computational demands arising from repeated evaluations of kernel integrals. MINPO, thus, generalizes beyond problem-specific formulations, providing a unified framework for systems governed by nonlocal operators.
【13】Do Foundational Audio Encoders Understand Music Structure?
标题：基础音频编码器了解音乐结构吗？
链接：https://arxiv.org/abs/2512.17209
摘要：在音乐信息检索（MIR）研究中，使用预训练的基础音频编码器（FAE）最近已成为一种趋势。在大量音乐和音频数据上预训练的FAE已被证明可以提高MIR任务的性能，例如音乐标记和自动音乐转录。然而，他们的使用音乐结构分析（MSA）仍然是探索不足。虽然有许多开源FAE模型可用，但只有一小部分已被用于MSA，并且学习方法，训练数据和模型上下文长度等因素对MSA性能的影响仍然不清楚。在这项研究中，我们对11种FAE进行了综合实验，以研究这些因素如何影响MSA性能。我们的研究结果表明，使用自监督学习的FAE与音乐数据上的掩蔽语言建模对MSA特别有效。这些发现为MSA的未来研究铺平了道路。
摘要：In music information retrieval (MIR) research, the use of pretrained foundational audio encoders (FAEs) has recently become a trend. FAEs pretrained on large amounts of music and audio data have been shown to improve performance on MIR tasks such as music tagging and automatic music transcription. However, their use for music structure analysis (MSA) remains underexplored. Although many open-source FAE models are available, only a small subset has been examined for MSA, and the impact of factors such as learning methods, training data, and model context length on MSA performance remains unclear. In this study, we conduct comprehensive experiments on 11 types of FAEs to investigate how these factors affect MSA performance. Our results demonstrate that FAEs using selfsupervised learning with masked language modeling on music data are particularly effective for MSA. These findings pave the way for future research in MSA.
【14】Digitizing Nepal's Written Heritage: A Comprehensive HTR Pipeline for Old Nepali Manuscripts
标题：尼泊尔文字遗产数字化：尼泊尔旧手稿的综合HTR管道
链接：https://arxiv.org/abs/2512.17111
备注：Under review
摘要：本文介绍了第一个端到端的管道手写文本识别（HTR）的旧尼泊尔语，一个历史上重要的，但低资源的语言。我们采用行级转录方法，并系统地探索编码器-解码器架构和以数据为中心的技术，以提高识别精度。我们的最佳模型实现了4.9%的字符错误率（CER）。此外，我们还实现和评估解码策略，并分析令牌级混淆，以更好地理解模型行为和错误模式。虽然我们用于评估的数据集是保密的，但我们发布了我们的训练代码，模型配置和评估脚本，以支持HTR对低资源历史脚本的进一步研究。
摘要：This paper presents the first end-to-end pipeline for Handwritten Text Recognition (HTR) for Old Nepali, a historically significant but low-resource language. We adopt a line-level transcription approach and systematically explore encoder-decoder architectures and data-centric techniques to improve recognition accuracy. Our best model achieves a Character Error Rate (CER) of 4.9\%. In addition, we implement and evaluate decoding strategies and analyze token-level confusions to better understand model behaviour and error patterns. While the dataset we used for evaluation is confidential, we release our training code, model configurations, and evaluation scripts to support further research in HTR for low-resource historical scripts.
【15】Atom: Efficient On-Device Video-Language Pipelines Through Modular Reuse
标题：Atom：通过模块化重用实现高效的设备上视频语言管道
链接：https://arxiv.org/abs/2512.17108
摘要：视频语言模型的最新进展使强大的应用程序，如视频检索，字幕和汇编。然而，由于冗余的模型加载和碎片化的执行，在移动设备上有效地执行这种多级流水线仍然具有挑战性。我们介绍原子，一个设备上的系统，重组视频语言管道快速和有效的执行。Atom将十亿个参数的模型分解为可重用的模块，如视觉编码器和语言解码器，并在字幕、推理和索引等子任务中重用它们。这种以重用为中心的设计消除了重复的模型加载，并支持并行执行，在不牺牲性能的情况下减少了端到端延迟。在普通智能手机上，Atom的执行速度比不重用的基线快27- 33%，性能只下降了一点点（检索时Recall@1为$\leq $2.3，字幕时CIDER为$\leq $1.5）。这些结果将Atom定位为在边缘设备上有效理解视频语言的实用、可扩展方法。
摘要：Recent advances in video-language models have enabled powerful applications like video retrieval, captioning, and assembly. However, executing such multi-stage pipelines efficiently on mobile devices remains challenging due to redundant model loads and fragmented execution. We introduce Atom, an on-device system that restructures video-language pipelines for fast and efficient execution. Atom decomposes a billion-parameter model into reusable modules, such as the visual encoder and language decoder, and reuses them across subtasks like captioning, reasoning, and indexing. This reuse-centric design eliminates repeated model loading and enables parallel execution, reducing end-to-end latency without sacrificing performance. On commodity smartphones, Atom achieves 27--33% faster execution compared to non-reuse baselines, with only marginal performance drop ($\leq$ 2.3 Recall@1 in retrieval, $\leq$ 1.5 CIDEr in captioning). These results position Atom as a practical, scalable approach for efficient video-language understanding on edge devices.
【16】Perturb Your Data: Paraphrase-Guided Training Data Watermarking
标题：扰乱您的数据：转述引导的训练数据水印
链接：https://arxiv.org/abs/2512.17075
备注：Accepted to AAAI 2026
摘要：训练数据检测对于执行版权和数据许可至关重要，因为大型语言模型（LLM）是在从互联网上抓取的大量文本语料库上训练的。我们提出了SPECTRA，水印的方法，使训练数据可靠地检测，即使它包括小于0.001%的训练语料库。SPECTRA的工作原理是使用LLM对文本进行释义，并根据单独的评分模型，根据每个释义的可能性分配一个分数。选择一个释义，使其得分密切匹配的原始文本，以避免引入任何分布的变化。为了测试可疑模型是否已经在水印数据上训练过，我们将其令牌概率与评分模型的令牌概率进行比较。我们证明了SPECTRA在检测用于训练的数据与未用于训练的数据时，达到了超过9个数量级的一致p值差距，这大于所有测试的基线。SPECTRA为数据所有者提供了一个可扩展的，发布前部署的水印，即使是大规模的LLM训练也能生存下来。
摘要：Training data detection is critical for enforcing copyright and data licensing, as Large Language Models (LLM) are trained on massive text corpora scraped from the internet. We present SPECTRA, a watermarking approach that makes training data reliably detectable even when it comprises less than 0.001% of the training corpus. SPECTRA works by paraphrasing text using an LLM and assigning a score based on how likely each paraphrase is, according to a separate scoring model. A paraphrase is chosen so that its score closely matches that of the original text, to avoid introducing any distribution shifts. To test whether a suspect model has been trained on the watermarked data, we compare its token probabilities against those of the scoring model. We demonstrate that SPECTRA achieves a consistent p-value gap of over nine orders of magnitude when detecting data used for training versus data not used for training, which is greater than all baselines tested. SPECTRA equips data owners with a scalable, deploy-before-release watermark that survives even large-scale LLM training.
【17】Universal consistency of the $k$-NN rule in metric spaces and Nagata dimension. III
标题：$k$-NN规则在度量空间和Nagata维度中的普遍一致性。III
链接：https://arxiv.org/abs/2512.17058
备注：12 pages, latex with ESAIM P&S macros
摘要：我们证明了最后一个剩余的蕴涵，允许对一个完全可分度量空间$X$要求以下条件的等价性： (1)k-最近邻分类器在X中是（弱）泛相容的，（2）对每一个局部有限Borel测度，强Lebesgue-Besicovitch微分性质在X中成立，（3）X在Nagata意义下是σ-有限维的. 等价式（2）$\iff$（3）由Preiss（1983）宣布，而蕴涵式（3）$\Rightarrow$（2）的详细证明出现在Assouad和Quentin de Gromard（2006）中。蕴涵（2）$\Rightarrow$（1）由Cérou和Guyader（2006）建立。我们证明了蕴涵（1）$\Rightarrow$（3）。该结果在该系列的第一篇文章中得到了证实（Collins，Kumari，Pestov 2020），在这里我们还纠正了第二篇文章中的错误主张（Kumari和Pestov 2024）。
摘要：We prove the last remaining implication allowing to claim the equivalence of the following conditions for a complete separable metric space $X$: (1) The $k$-nearest neighbour classifier is (weakly) universally consistent in $X$, (2) The strong Lebesgue--Besicovitch differentiation property holds in $X$ for every locally finite Borel measure, (3) $X$ is sigma-finite dimensional in the sense of Nagata. The equivalence (2)$\iff$(3) was announced by Preiss (1983), while a detailed proof of the implication (3)$\Rightarrow$(2) has appeared in Assouad and Quentin de Gromard (2006). The implication (2)$\Rightarrow$(1) was established by Cérou and Guyader (2006). We prove the implication (1)$\Rightarrow$(3). The result was conjectured in the first article in the series (Collins, Kumari, Pestov 2020), and here we also correct a wrong claim made in the second article (Kumari and Pestov 2024).
【18】Dynamic Tool Dependency Retrieval for Efficient Function Calling
标题：动态工具依赖性检索以实现高效的函数调用
链接：https://arxiv.org/abs/2512.17052
备注：18 pages, 5 figures, 6 tables
摘要：由大型语言模型（LLM）支持的函数调用代理选择外部工具来自动化复杂的任务。设备上代理通常使用检索模块来选择相关工具，从而提高性能并缩短上下文长度。然而，现有的检索方法依赖于静态和有限的输入，未能捕获多步骤的工具依赖性和不断变化的任务上下文。这种限制通常会引入不相关的工具，误导代理，降低效率和准确性。我们提出了动态工具依赖检索（DTDR），一个轻量级的检索方法，条件的初始查询和不断变化的执行上下文。DTDR从函数调用演示中对工具依赖性进行建模，从而在计划展开时实现自适应检索。我们对DTDR进行了跨多个数据集和LLM主干的最先进的检索方法的基准测试，评估了检索精度，下游任务准确性和计算效率。此外，我们探索策略，将检索到的工具到提示。我们的研究结果表明，动态工具检索提高函数调用成功率之间的23\%$和104\%$相比，国家的最先进的静态检索。
摘要 ：Function calling agents powered by Large Language Models (LLMs) select external tools to automate complex tasks. On-device agents typically use a retrieval module to select relevant tools, improving performance and reducing context length. However, existing retrieval methods rely on static and limited inputs, failing to capture multi-step tool dependencies and evolving task context. This limitation often introduces irrelevant tools that mislead the agent, degrading efficiency and accuracy. We propose Dynamic Tool Dependency Retrieval (DTDR), a lightweight retrieval method that conditions on both the initial query and the evolving execution context. DTDR models tool dependencies from function calling demonstrations, enabling adaptive retrieval as plans unfold. We benchmark DTDR against state-of-the-art retrieval methods across multiple datasets and LLM backbones, evaluating retrieval precision, downstream task accuracy, and computational efficiency. Additionally, we explore strategies to integrate retrieved tools into prompts. Our results show that dynamic tool retrieval improves function calling success rates between $23\%$ and $104\%$ compared to state-of-the-art static retrievers.
【19】PAACE: A Plan-Aware Automated Agent Context Engineering Framework
标题：PAACE：一个计划感知的自动化代理上下文工程框架
链接：https://arxiv.org/abs/2512.16970
摘要：大型语言模型（LLM）代理越来越多地部署在复杂的多步骤工作流中，涉及规划，工具使用，反思以及与外部知识系统的交互。这些工作流生成快速扩展的上下文，必须对其进行策划、转换和压缩，以保持保真度，避免注意力稀释，并降低推理成本。以前的工作总结和查询感知压缩在很大程度上忽略了多步骤，计划感知的性质，代理推理。在这项工作中，我们引入了PAACE（计划感知自动上下文工程），这是一个统一的框架，用于通过下一个k任务相关性建模，计划结构分析，指令协同细化和功能保留压缩来优化LLM代理的演变状态。PAACE包括（1）PAACE-Syn，一个带有逐步压缩监督注释的合成代理工作流的大规模生成器，以及（2）PAACE-FT，一系列经过成功教师演示培训的蒸馏，计划感知压缩器。在长期基准测试（AppWorld，WEBBench和8-Objective QA）上的实验表明，PAACE在大幅降低上下文负载的同时，始终提高了代理的正确性。在AppWorld上，PAACE实现了比所有基线更高的准确性，同时降低了峰值上下文和累积依赖性。PAACE在测试台和多跳QA上提高了准确性和F1，实现了更少的步骤、更低的峰值令牌和更低的注意力依赖性。经过提炼的PAACE-FT保留了教师97%的性能，同时将推理成本降低了一个数量级以上，从而实现了使用紧凑模型的计划感知压缩的实际部署。
摘要：Large Language Model (LLM) agents are increasingly deployed in complex, multi-step workflows involving planning, tool use, reflection, and interaction with external knowledge systems. These workflows generate rapidly expanding contexts that must be curated, transformed, and compressed to maintain fidelity, avoid attention dilution, and reduce inference cost. Prior work on summarization and query-aware compression largely ignores the multi-step, plan-aware nature of agentic reasoning. In this work, we introduce PAACE (Plan-Aware Automated Context Engineering), a unified framework for optimizing the evolving state of LLM agents through next-k-task relevance modeling, plan-structure analysis, instruction co-refinement, and function-preserving compression. PAACE comprises (1) PAACE-Syn, a large-scale generator of synthetic agent workflows annotated with stepwise compression supervision, and (2) PAACE-FT, a family of distilled, plan-aware compressors trained from successful teacher demonstrations. Experiments on long-horizon benchmarks (AppWorld, OfficeBench, and 8-Objective QA) demonstrate that PAACE consistently improves agent correctness while substantially reducing context load. On AppWorld, PAACE achieves higher accuracy than all baselines while lowering peak context and cumulative dependency. On OfficeBench and multi-hop QA, PAACE improves both accuracy and F1, achieving fewer steps, lower peak tokens, and reduced attention dependency. Distilled PAACE-FT retains 97 percent of the teacher's performance while reducing inference cost by over an order of magnitude, enabling practical deployment of plan-aware compression with compact models.
【20】SpIDER: Spatially Informed Dense Embedding Retrieval for Software Issue Localization
标题：SpIDER：用于软件问题本地化的空间知情密集嵌入检索
链接：https://arxiv.org/abs/2512.16956
备注：Initial preprint
摘要：检索代码单元（例如，文件、类、函数）在语义上与来自大型代码库的给定用户查询、错误报告或特征请求相关是基于LLM的编码代理的基本挑战。解析方法通常采用稀疏检索方法（如BM 25）或密集嵌入策略来识别相关单元。虽然基于嵌入的方法可以大大优于BM 25，但它们通常缺乏对代码库的探索，并且没有充分利用其底层图形结构。为了解决这个问题，我们提出了SpIDER（空间信息密集嵌入检索），这是一种增强的密集检索方法，它结合了基于LLM的推理，通过基于图形的代码库探索获得辅助上下文。实证结果表明，SpIDER在几种编程语言中一致地提高了密集检索性能。
摘要：Retrieving code units (e.g., files, classes, functions) that are semantically relevant to a given user query, bug report, or feature request from large codebases is a fundamental challenge for LLM-based coding agents. Agentic approaches typically employ sparse retrieval methods like BM25 or dense embedding strategies to identify relevant units. While embedding-based approaches can outperform BM25 by large margins, they often lack exploration of the codebase and underutilize its underlying graph structure. To address this, we propose SpIDER (Spatially Informed Dense Embedding Retrieval), an enhanced dense retrieval approach that incorporates LLM-based reasoning over auxiliary context obtained through graph-based exploration of the codebase. Empirical results show that SpIDER consistently improves dense retrieval performance across several programming languages.
【21】BIONIX: A Wireless, Low-Cost Prosthetic Arm with Dual-Signal EEG and EMG Control
标题：BIONIX：一款具有双信号脑电和EMG控制的无线、低成本假肢手臂
链接：https://arxiv.org/abs/2512.16929
备注：12 pages, 8 figures
摘要：负担得起的上肢假肢往往缺乏直观的控制系统，限制了截肢者在低资源环境中的功能和可及性。该项目提出了一种低成本的双模式神经肌肉控制系统，该系统集成了脑电图（EEG）和肌电图（EMG），以实现对假肢的实时多自由度控制。EEG信号使用NeuroSky MindWave Mobile 2采集，并通过ThinkGear蓝牙数据包传输到运行轻量级分类模型的ESP 32微控制器。该模型使用具有低通滤波的6帧滑动窗口在1500秒记录的EEG数据上进行训练，排除不良信号样本，并使用70/20/10训练-验证-测试分割。分类器检测强眨眼事件，其在打开和闭合状态之间切换手。EMG信号使用MyoWare 2.0传感器和SparkFun无线屏蔽获取，并传输到第二个ESP 32，它执行基于阈值的检测。三个激活带（其余：0--T1;分机号：T1-T2;收缩：大于T2）能够实现直观的肘关节控制，仅在运动类别中的八个连续帧之后才触发运动，以提高稳定性。EEG控制的ESP 32驱动四个手指伺服，而EMG控制的ESP 32驱动两个肘部伺服。使用低成本材料（总成本约240美元）构建了一个功能原型，其中大部分费用归因于商业EEG耳机。未来的工作包括过渡到3D打印的底盘，集成自回归模型以减少EMG延迟，以及升级伺服扭矩以提高负载能力和握力。该系统展示了一种可行的途径，以低成本，生物直观的假肢控制，适用于服务不足和全球健康的应用。
摘要：Affordable upper-limb prostheses often lack intuitive control systems, limiting functionality and accessibility for amputees in low-resource settings. This project presents a low-cost, dual-mode neuro-muscular control system integrating electroencephalography (EEG) and electromyography (EMG) to enable real-time, multi-degree-of-freedom control of a prosthetic arm. EEG signals are acquired using the NeuroSky MindWave Mobile 2 and transmitted via ThinkGear Bluetooth packets to an ESP32 microcontroller running a lightweight classification model. The model was trained on 1500 seconds of recorded EEG data using a 6-frame sliding window with low-pass filtering, excluding poor-signal samples and using a 70/20/10 training--validation--test split. The classifier detects strong blink events, which toggle the hand between open and closed states. EMG signals are acquired using a MyoWare 2.0 sensor and SparkFun wireless shield and transmitted to a second ESP32, which performs threshold-based detection. Three activation bands (rest: 0--T1; extension: T1--T2; contraction: greater than T2) enable intuitive elbow control, with movement triggered only after eight consecutive frames in a movement class to improve stability. The EEG-controlled ESP32 actuates four finger servos, while the EMG-controlled ESP32 drives two elbow servos. A functional prototype was constructed using low-cost materials (total cost approximately 240 dollars), with most expense attributed to the commercial EEG headset. Future work includes transitioning to a 3D-printed chassis, integrating auto-regressive models to reduce EMG latency, and upgrading servo torque for improved load capacity and grip strength. This system demonstrates a feasible pathway to low-cost, biologically intuitive prosthetic control suitable for underserved and global health applications.
【22】Dion2: A Simple Method to Shrink Matrix in Muon
标题
：Dion 2：在μ子中缩小矩阵的简单方法
链接：https://arxiv.org/abs/2512.16928
备注：https://github.com/microsoft/dion/
摘要：μ子优化器具有强大的经验性能和理论基础。然而，其正交归一化步骤的超线性成本引入了随着规模而增加的开销。为了减轻这种成本，一些作品试图减少进入正交归一化步骤的矩阵的大小。我们介绍Dion2，一个更简单的方法，用于收缩矩阵中涉及μ子的计算相比，以前的方法。在高级别上，Dion2在每次迭代时选择一部分行或列，并仅对这些行或列进行正交归一化。这种采样过程使得更新稀疏，降低了计算和通信成本，从而提高了μ子的可扩展性。
摘要：The Muon optimizer enjoys strong empirical performance and theoretical grounding. However, the super-linear cost of its orthonormalization step introduces increasing overhead with scale. To alleviate this cost, several works have attempted to reduce the size of the matrix entering the orthonormalization step. We introduce Dion2, a much simpler method for shrinking the matrix involved in Muon's computation compared to prior approaches. At a high level, Dion2 selects a fraction of rows or columns at each iteration and orthonormalizes only those. This sampling procedure makes the update sparse, reducing both computation and communication costs which in turn improves the scalability of Muon.
【23】Optimizing Text Search: A Novel Pattern Matching Algorithm Based on Ukkonen's Approach
标题：优化文本搜索：一种基于Ukkonen方法的模式匹配算法
链接：https://arxiv.org/abs/2512.16927
备注：5 pages, 13 figures
摘要：在计算机科学领域，文本搜索算法的效率对于处理自然语言处理和生物信息学等领域的大量数据至关重要。传统的方法，如朴素搜索、KMP和Boyer-Moore，虽然是基础性的，但在处理现代数据集的复杂性和规模方面往往不足，如路透社语料库和人类基因组序列。这项研究严格调查了文本搜索算法，重点是通过Splitting和Ukkonen算法等方法优化后缀树，分析了包括路透社语料库和人类基因组在内的数据集。一种新的优化相结合Ukkonen的算法与一种新的搜索技术，显示线性的时间和空间效率，优于传统的方法，如朴素搜索，KMP，和Boyer-Moore。实证测试证实了理论上的优势，突出了优化的后缀树在基因组序列模式识别等任务中的有效性，实现了100%的准确率。这项研究不仅推进了文本搜索算法的学术知识，而且由于其优越的资源效率和可靠性，在自然语言处理和生物信息学等领域也表现出了重要的实用性。
摘要：In the realm of computer science, the efficiency of text-search algorithms is crucial for processing vast amounts of data in areas such as natural language processing and bioinformatics. Traditional methods like Naive Search, KMP, and Boyer-Moore, while foundational, often fall short in handling the complexities and scale of modern datasets, such as the Reuters corpus and human genomic sequences. This study rigorously investigates text-search algorithms, focusing on optimizing Suffix Trees through methods like Splitting and Ukkonen's Algorithm, analyzed on datasets including the Reuters corpus and human genomes. A novel optimization combining Ukkonen's Algorithm with a new search technique is introduced, showing linear time and space efficiencies, outperforming traditional methods like Naive Search, KMP, and Boyer-Moore. Empirical tests confirm the theoretical advantages, highlighting the optimized Suffix Tree's effectiveness in tasks like pattern recognition in genomic sequences, achieving 100% accuracy. This research not only advances academic knowledge in text-search algorithms but also demonstrates significant practical utility in fields like natural language processing and bioinformatics, due to its superior resource efficiency and reliability.
【24】Domain-Aware Quantum Circuit for QML
标题：QML领域感知量子电路
链接：https://arxiv.org/abs/2512.17800
摘要：设计具有表达性、可训练性和对硬件噪声鲁棒性的参数化量子电路（PQC）是噪声中间尺度量子（NISQ）设备上量子机器学习（QML）的核心挑战。我们提出了一个域感知量子电路（DAQC），利用图像先验知识，通过非重叠的DCT风格的锯齿形窗口来指导局部保持编码和纠缠。该设计采用交织的编码-纠缠-训练循环，其中纠缠被应用于托管相邻像素的量子比特之间，与设备连接性对准。这种分阶段的局部保持信息流在没有深度全局混合的情况下扩展了有效的感受野，从而能够有效地使用有限的深度和量子位。该设计将表征能力集中在短程相关性上，减少了长程双量子比特操作，并鼓励稳定优化，从而减轻了深度诱导和全局纠缠的贫瘠高原效应。我们在MNIST、FashionMNIST和MononiaMNIST数据集上评估DAQC。在量子硬件上，DAQC实现了与强经典基线（例如，ResNet-18/50，DenseNet-121，EfficientNet-B 0），并大大优于量子电路搜索（QCS）基线。据我们所知，DAQC使用仅具有线性经典读出（没有深度经典主干）的量子特征提取器，目前在基于QML的图像分类任务的真实量子硬件上实现了最佳性能。代码和预训练模型可在https://github.com/gurinder-hub/DAQC上获得。
摘要：Designing parameterized quantum circuits (PQCs) that are expressive, trainable, and robust to hardware noise is a central challenge for quantum machine learning (QML) on noisy intermediate-scale quantum (NISQ) devices. We present a Domain-Aware Quantum Circuit (DAQC) that leverages image priors to guide locality-preserving encoding and entanglement via non-overlapping DCT-style zigzag windows. The design employs interleaved encode-entangle-train cycles, where entanglement is applied among qubits hosting neighboring pixels, aligned to device connectivity. This staged, locality-preserving information flow expands the effective receptive field without deep global mixing, enabling efficient use of limited depth and qubits. The design concentrates representational capacity on short-range correlations, reduces long-range two-qubit operations, and encourages stable optimization, thereby mitigating depth-induced and globally entangled barren-plateau effects. We evaluate DAQC on MNIST, FashionMNIST, and PneumoniaMNIST datasets. On quantum hardware, DAQC achieves performance competitive with strong classical baselines (e.g., ResNet-18/50, DenseNet-121, EfficientNet-B0) and substantially outperforming Quantum Circuit Search (QCS) baselines. To the best of our knowledge, DAQC, which uses a quantum feature extractor with only a linear classical readout (no deep classical backbone), currently achieves the best reported performance on real quantum hardware for QML-based image classification tasks. Code and pretrained models are available at: https://github.com/gurinder-hub/DAQC.
【25】Alternating Direction Method of Multipliers for Nonlinear Matrix Decompositions
标题：非线性矩阵分解的乘子交替方向法
链接：https://arxiv.org/abs/2512.17473
备注：14 pages, 6 figures. Code available from https://gitlab.com/Atharva05/admm-for-nmd
摘要：提出了一种基于交替方向乘子法（ADMM）的非线性矩阵分解（NMD）算法。给定输入矩阵$X \in \mathbb{R}^{m \times n}$和分解秩$r \ll \min（m，n）$，NMD寻找矩阵$W \in \mathbb{R}^{m \times r}$和$H \in \mathbb{R}^{r \times n}$，使得$X \approx f（WH）$，其中$f$是逐元素非线性函数。我们在几个代表性的非线性模型上评估我们的方法：整流线性单元激活$f（x）= \max（0，x）$，适用于非负稀疏数据近似，分量平方$f（x）= x^2$，适用于概率电路表示，以及MinMax变换$f（x）= \min（b，\max（a，x））$，与推荐系统相关。所提出的框架灵活地支持不同的损失函数，包括最小二乘，$\ell_1 $范数，和Kullback-Leibler分歧，并可以很容易地扩展到其他非线性和度量。我们说明了该方法在现实世界数据集上的适用性，效率和适应性，突出了其广泛应用的潜力。
摘要 ：We present an algorithm based on the alternating direction method of multipliers (ADMM) for solving nonlinear matrix decompositions (NMD). Given an input matrix $X \in \mathbb{R}^{m \times n}$ and a factorization rank $r \ll \min(m, n)$, NMD seeks matrices $W \in \mathbb{R}^{m \times r}$ and $H \in \mathbb{R}^{r \times n}$ such that $X \approx f(WH)$, where $f$ is an element-wise nonlinear function. We evaluate our method on several representative nonlinear models: the rectified linear unit activation $f(x) = \max(0, x)$, suitable for nonnegative sparse data approximation, the component-wise square $f(x) = x^2$, applicable to probabilistic circuit representation, and the MinMax transform $f(x) = \min(b, \max(a, x))$, relevant for recommender systems. The proposed framework flexibly supports diverse loss functions, including least squares, $\ell_1$ norm, and the Kullback-Leibler divergence, and can be readily extended to other nonlinearities and metrics. We illustrate the applicability, efficiency, and adaptability of the approach on real-world datasets, highlighting its potential for a broad range of applications.
【26】Perfect reconstruction of sparse signals using nonconvexity control and one-step RSB message passing
标题：使用非凸控制和一步RSB消息传递完美重建稀疏信号
链接：https://arxiv.org/abs/2512.17426
备注：49 pages, 10 figures
摘要：我们认为稀疏信号重建通过最小化的平滑剪切绝对偏差（SCAD）的惩罚，并开发一步复制对称性破坏（1 RSB）的近似消息传递（AMP）的扩展，称为1 RSB-AMP。从1 RSB信念传播公式出发，我们推导出1 RSB-AMP的更新规则以及相应的状态演化（1 RSB-SE）方程。一个详细的比较表明，1 RSB-AMP和1 RSB-SE同意非常好的宏观水平，即使在参数区域复制对称（RS）AMP，称为RS-AMP，发散和1 RSB描述本身预计不会是精确的。1 RSB-SE的定点分析揭示了一个由成功、失败和发散阶段组成的相图，与RS情况一样。然而，发散区域的边界现在取决于Parisi参数，由于1 RSB animals，我们提出了一个新的标准-最小化发散区域的大小-而不是传统的零复杂性条件，以确定其值。结合这个标准与非凸性控制（NCC）协议在以前的RS研究中提出的改进的算法限制的完美重建相比，RS-AMP。1 RSB-SE的数值解和使用1 RSB-AMP的实验证实，这种改进的限制在实践中是实现的，尽管增益是适度的，并且仍然略低于贝叶斯最优阈值。我们还报告了热力学量的行为-重叠，自由熵，复杂性和非自平均磁化率-在这个问题中的1 RSB阶段的特点。
摘要：We consider sparse signal reconstruction via minimization of the smoothly clipped absolute deviation (SCAD) penalty, and develop one-step replica-symmetry-breaking (1RSB) extensions of approximate message passing (AMP), termed 1RSB-AMP. Starting from the 1RSB formulation of belief propagation, we derive explicit update rules of 1RSB-AMP together with the corresponding state evolution (1RSB-SE) equations. A detailed comparison shows that 1RSB-AMP and 1RSB-SE agree remarkably well at the macroscopic level, even in parameter regions where replica-symmetric (RS) AMP, termed RS-AMP, diverges and where the 1RSB description itself is not expected to be thermodynamically exact. Fixed-point analysis of 1RSB-SE reveals a phase diagram consisting of success, failure, and diverging phases, as in the RS case. However, the diverging-region boundary now depends on the Parisi parameter due to the 1RSB ansatz, and we propose a new criterion -- minimizing the size of the diverging region -- rather than the conventional zero-complexity condition, to determine its value. Combining this criterion with the nonconvexity-control (NCC) protocol proposed in a previous RS study improves the algorithmic limit of perfect reconstruction compared with RS-AMP. Numerical solutions of 1RSB-SE and experiments with 1RSB-AMP confirm that this improved limit is achieved in practice, though the gain is modest and remains slightly inferior to the Bayes-optimal threshold. We also report the behavior of thermodynamic quantities -- overlaps, free entropy, complexity, and the non-self-averaging susceptibility -- that characterize the 1RSB phase in this problem.
机器翻译由腾讯交互翻译提供，仅供参考
点击“阅读原文”获取带摘要的学术速递
Python社区是高质量的Python/Django开发社区
本文地址：http://www.python88.com/topic/190814
本文地址：http://www.python88.com/topic/190814