|
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154篇
大模型相关(23篇)
【1】Measuring Faithfulness Depends on How You Measure: Classifier Sensitivity in LLM Chain-of-Thought Evaluation
标题：衡量忠诚度取决于如何衡量：LLM思想链评估中的分类敏感性
链接：https://arxiv.org/abs/2603.20172
备注：14 pages, 4 figures, 5 tables
摘要：最近关于思想链（CoT）忠诚度的研究报告了单个聚合数字（例如，DeepSeek-R1在39%的情况下承认提示，这意味着忠诚是模型的一个客观，可衡量的属性。本文证明，它不是。三个分类器（仅regex检测器，两阶段regex加LLM管道和独立的Claude Sonnet 4判断）应用于10，276个受影响的推理轨迹，这些推理轨迹来自12个开放权重模型，跨越9个家族和7 B到1 T参数。在相同的数据上，这些分类器分别产生74.4%、82.6%和69.7%的总体忠实率，95%的置信区间不重叠。每个模型的差距范围为2.6至30.6个百分点;所有差异均具有统计学显著性（McNemar检验，p < 0.001）。分歧是系统性的，而不是随机的：科恩的kappa测量的分类器间的一致性范围从0.06（“轻微”）的奉承暗示到0.42（“中度”）的分级提示，不对称是明显的：对于奉承，883例被管道分类为忠实，但不忠实的十四行诗法官，而只有2个去相反的方向。分类器的选择也可以逆转模型的排名：Qwen3.5- 27 B在管道下排名第一，但在Sonnet判断下排名第七; OLMo-3.1- 32 B则相反，从第9位上升到第3位。其根本原因是，不同的分类操作相关的忠实性结构在不同水平的严格性（词汇提及与认知依赖），这些结构产生不同的测量相同的行为。这些结果表明，在使用不同分类器的研究中，无法对已发表的忠诚度数字进行有意义的比较，未来的评估应报告多种分类方法的敏感性范围，而不是单点估计。
摘要：Recent work on chain-of-thought (CoT) faithfulness reports single aggregate numbers (e.g., DeepSeek-R1 acknowledges hints 39% of the time), implying that faithfulness is an objective, measurable property of a model. This paper demonstrates that it is not. Three classifiers (a regex-only detector, a two-stage regex-plus-LLM pipeline, and an independent Claude Sonnet 4 judge) are applied to 10,276 influenced reasoning traces from 12 open-weight models spanning 9 families and 7B to 1T parameters. On identical data, these classifiers produce overall faithfulness rates of 74.4%, 82.6%, and 69.7%, respectively, with non-overlapping 95% confidence intervals. Per-model gaps range from 2.6 to 30.6 percentage points; all are statistically significant (McNemar's test, p < 0.001). The disagreements are systematic, not random: inter-classifier agreement measured by Cohen's kappa ranges from 0.06 ("slight") for sycophancy hints to 0.42 ("moderate") for grader hints, and the asymmetry is pronounced: for sycophancy, 883 cases are classified as faithful by the pipeline but unfaithful by the Sonnet judge, while only 2 go the other direction. Classifier choice can also reverse model rankings: Qwen3.5-27B ranks 1st under the pipeline but 7th under the Sonnet judge; OLMo-3.1-32B moves in the opposite direction, from 9th to 3rd. The root cause is that different classifiers operationalize related faithfulness constructs at different levels of stringency (lexical mention versus epistemic dependence), and these constructs yield divergent measurements on the same behavior. These results demonstrate that published faithfulness numbers cannot be meaningfully compared across studies that use different classifiers, and that future evaluations should report sensitivity ranges across multiple classification methodologies rather than single point estimates.
【2】Semantic Token Clustering for Efficient Uncertainty Quantification in Large Language Models
标题：语义令牌聚集用于大型语言模型中的高效不确定性量化
链接：https://arxiv.org/abs/2603.20161
备注：EACL 2026
摘要：大型语言模型（LLM）在不同的任务中表现出了卓越的能力。然而，他们输出的真实性并不能得到保证，他们过度自信的倾向进一步限制了可靠性。不确定性量化提供了一种有前途的方法来识别潜在的不可靠的输出，但大多数现有的方法依赖于重复采样或辅助模型，引入大量的计算开销。为了解决这些限制，我们提出了语义令牌聚类（STC），一种有效的不确定性量化方法，利用LLM中固有的语义信息编码。具体来说，我们使用嵌入聚类和前缀匹配将令牌分组为语义一致的簇，并基于在相应的语义簇上聚合的概率质量来量化不确定性。我们的方法只需要一个单一的代，不依赖于辅助模型。实验结果表明，STC实现的性能与国家的最先进的基线，同时大大减少了计算开销。
摘要：Large language models (LLMs) have demonstrated remarkable capabilities across diverse tasks. However, the truthfulness of their outputs is not guaranteed, and their tendency toward overconfidence further limits reliability. Uncertainty quantification offers a promising way to identify potentially unreliable outputs, but most existing methods rely on repeated sampling or auxiliary models, introducing substantial computational overhead. To address these limitations, we propose Semantic Token Clustering (STC), an efficient uncertainty quantification method that leverages the semantic information inherently encoded in LLMs. Specifically, we group tokens into semantically consistent clusters using embedding clustering and prefix matching, and quantify uncertainty based on the probability mass aggregated over the corresponding semantic cluster. Our approach requires only a single generation and does not depend on auxiliary models. Experimental results show that STC achieves performance comparable to state-of-the-art baselines while substantially reducing computational overhead.
【3】The $\mathbf{Y}$-Combinator for LLMs: Solving Long-Context Rot with $λ$-Calculus
链接：https://arxiv.org/abs/2603.20105
摘要：LLM越来越多地用作通用推理机，但长输入仍然由固定的上下文窗口检查。递归语言模型（RLM）通过外部化提示和递归解决子问题来解决这个问题。然而，现有的RLM依赖于一个开放式的读取-评估-打印循环（REPL），其中模型生成任意控制代码，使得执行难以验证，预测和分析。 我们引入了$λ$-RLM，一个长上下文推理的框架，它用基于$λ$-演算的类型化函数运行时代替了自由形式的递归代码生成。它执行一个紧凑的预验证组合子库，并只使用有界叶子子问题的神经推理，将递归推理转化为具有显式控制流的结构化功能程序。我们表明，$λ$-RLM承认正式保证缺乏标准RLM，包括终止，封闭形式的成本界，控制精度缩放递归深度，和一个最佳的分区规则下的一个简单的成本模型。从经验上看，在四个长上下文推理任务和九个基本模型中，$λ$-RLM在36个模型-任务比较中的29个中优于标准RLM，在模型层中将平均准确率提高了+21.9个点，并将延迟降低了4.1倍。这些结果表明，类型化的符号控制产生了一个更可靠和有效的基础，长上下文推理比开放式递归代码生成。$λ$-RLM的完整实现在https://github.com/lambda-calculus-LLM/lambda-RLM上对社区开放。
摘要 ：LLMs are increasingly used as general-purpose reasoners, but long inputs remain bottlenecked by a fixed context window. Recursive Language Models (RLMs) address this by externalising the prompt and recursively solving subproblems. Yet existing RLMs depend on an open-ended read-eval-print loop (REPL) in which the model generates arbitrary control code, making execution difficult to verify, predict, and analyse. We introduce $λ$-RLM, a framework for long-context reasoning that replaces free-form recursive code generation with a typed functional runtime grounded in $λ$-calculus. It executes a compact library of pre-verified combinators and uses neural inference only on bounded leaf subproblems, turning recursive reasoning into a structured functional program with explicit control flow. We show that $λ$-RLM admits formal guarantees absent from standard RLMs, including termination, closed-form cost bounds, controlled accuracy scaling with recursion depth, and an optimal partition rule under a simple cost model. Empirically, across four long-context reasoning tasks and nine base models, $λ$-RLM outperforms standard RLM in 29 of 36 model-task comparisons, improves average accuracy by up to +21.9 points across model tiers, and reduces latency by up to 4.1x. These results show that typed symbolic control yields a more reliable and efficient foundation for long-context reasoning than open-ended recursive code generation. The complete implementation of $λ$-RLM, is open-sourced for the community at: https://github.com/lambda-calculus-LLM/lambda-RLM.
【4】Breaking the Capability Ceiling of LLM Post-Training by Reintroducing Markov States
标题：通过重新引入马尔科夫状态打破LLM训练后的能力上限
链接：https://arxiv.org/abs/2603.19987
摘要：强化学习（RL）已经成为后训练和对齐大型语言模型（LLM）的标准范例，但最近的证据表明它面临着持久的“能力上限”：与发现新策略的经典RL系统不同，LLM的RL通常只是对已经潜伏在预训练权重中的模式的细化。在这项工作中，我们确定了一个基本的结构瓶颈：虽然经典的RL依赖于紧凑的，信息丰富的马尔可夫状态，但当前的LLM后训练公式与不断扩展的动作历史有关。 我们重新审视了一个经典的原则，长期以来一直是RL的核心，但在LLM后训练中却没有：显式马尔可夫状态。从理论上讲，我们提供了严格的保证，证明利用估计的马尔可夫状态可以显着降低样本的复杂性。从经验上讲，我们表明，在一系列复杂的逻辑谜题中，引入马尔可夫状态始终打破了标准RL后训练的性能边界。我们的研究结果表明，超越“历史即状态”建模，支持结构化马尔可夫表示，对于解锁生成式AI中的开放式发现和真正的新推理能力至关重要。
摘要：Reinforcement learning (RL) has become a standard paradigm for post-training and aligning Large Language Models (LLMs), yet recent evidence suggests it faces a persistent "capability ceiling": unlike classical RL systems that discover novel strategies, RL for LLMs often acts as a mere refiner of patterns already latent in pre-trained weights. In this work, we identify a fundamental structural bottleneck: while classical RL relies on compact, informative Markov states, current LLM post-training formulations are tethered to an ever-expanding history of actions. We revisit a classical principle long central to RL yet absent from LLM post-training: explicit Markov states. Theoretically, we provide rigorous guarantees demonstrating that leveraging estimated Markov states can significantly reduce sample complexity. Empirically, we show that introducing Markov states consistently breaks the performance boundaries of standard RL post-training across a suite of complex logic puzzles. Our findings suggest that moving beyond "history-as-state" modeling in favor of structured Markovian representations is essential for unlocking open-ended discovery and genuinely new reasoning capabilities in Generative AI.
【5】Memori: A Persistent Memory Layer for Efficient, Context-Aware LLM Agents
标题：Memori：用于高效、上下文感知的LLM代理的持久内存层
链接：https://arxiv.org/abs/2603.19935
备注：9 pages; 2 figures; white paper
摘要：随着大型语言模型（LLM）发展成为自治代理，API层的持久内存对于实现跨LLM和多会话交互的上下文感知行为至关重要。现有的方法迫使供应商锁定，并依赖于将大量原始对话注入提示，导致令牌成本高，性能下降。 我们介绍Memori，一个与LLM无关的持久内存层，它将内存视为数据结构问题。其高级增强管道将非结构化对话转换为紧凑的语义三元组和对话摘要，从而实现精确的检索和连贯的推理。 根据LoCoMo基准测试进行评估，Memori的准确率为81.95%，优于现有的内存系统，同时每个查询仅使用1，294个令牌（约为完整上下文的5%）。这导致了成本的大幅降低，包括比竞争方法少67%的令牌，与全上下文方法相比节省了20倍以上。 这些结果表明，LLM代理中的有效内存取决于结构化表示，而不是更大的上下文窗口，从而实现可扩展和具有成本效益的部署。
摘要：As large language models (LLMs) evolve into autonomous agents, persistent memory at the API layer is essential for enabling context-aware behavior across LLMs and multi-session interactions. Existing approaches force vendor lock-in and rely on injecting large volumes of raw conversation into prompts, leading to high token costs and degraded performance. We introduce Memori, an LLM-agnostic persistent memory layer that treats memory as a data structuring problem. Its Advanced Augmentation pipeline converts unstructured dialogue into compact semantic triples and conversation summaries, enabling precise retrieval and coherent reasoning. Evaluated on the LoCoMo benchmark, Memori achieves 81.95% accuracy, outperforming existing memory systems while using only 1,294 tokens per query (~5% of full context). This results in substantial cost reductions, including 67% fewer tokens than competing approaches and over 20x savings compared to full-context methods. These results show that effective memory in LLM agents depends on structured representations instead of larger context windows, enabling scalable and cost-efficient deployment.
【6】FedPDPO: Federated Personalized Direct Preference Optimization for Large Language Model Alignment
标题：FedPDPO：针对大型语言模型对齐的联合个性化直接偏好优化
链接：https://arxiv.org/abs/2603.19741
备注：under review
摘要：在联邦学习（FL）中，将大型语言模型（LLM）与人类偏好对齐是一项挑战，因为数据是分散的、隐私敏感的和高度非IID偏好的。直接偏好优化（DPO）为带有人类反馈的强化学习（RLHF）提供了一种有效的替代方案，但其在FL中的直接应用在非IID数据下性能严重下降，并且隐式奖励的推广有限。为了弥合这一差距，我们提出了FedPDPO（联邦个性化直接偏好优化），一个个性化的联邦框架LLM的偏好对齐。它采用了一种参数高效的微调架构，每个客户端都维护一个冻结的预训练LLM骨干，并使用低秩自适应（LoRA）适配器进行扩展，从而实现通信高效的聚合。为了解决非IID异构性，我们设计了（1）全球共享的LoRA适配器，带有个性化的客户特定LLM头。此外，我们还引入了（2）一种个性化的DPO训练策略，该策略具有特定于客户端的显式奖励头部，以补充隐式奖励并进一步缓解非IID异构性，以及（3）一个瓶颈适配器，以平衡全局和局部特征。我们提供理论分析，建立概率基础和可靠性。在多个偏好数据集上进行的大量实验证明了最先进的性能，在联合域内和跨域设置中实现了高达4.80%的平均准确率提高。
摘要：Aligning large language models (LLMs) with human preferences in federated learning (FL) is challenging due to decentralized, privacy-sensitive, and highly non-IID preference data. Direct Preference Optimization (DPO) offers an efficient alternative to reinforcement learning with human feedback (RLHF), but its direct application in FL suffers from severe performance degradation under non-IID data and limited generalization of implicit rewards. To bridge this gap, we propose FedPDPO (Federated Personalized Direct Preference Optimization), a personalized federated framework for preference alignment of LLMs. It adopts a parameter-efficient fine-tuning architecture where each client maintains a frozen pretrained LLM backbone augmented with a Low-Rank Adaptation (LoRA) adapter, enabling communication-efficient aggregation. To address non-IID heterogeneity, we devise (1) the globally shared LoRA adapter with the personalized client-specific LLM head. Moreover, we introduce (2) a personalized DPO training strategy with a client-specific explicit reward head to complement implicit rewards and further alleviate non-IID heterogeneity, and (3) a bottleneck adapter to balance global and local features. We provide theoretical analysis establishing the probabilistic foundation and soundness. Extensive experiments on multiple preference datasets demonstrate state-of-the-art performance, achieving up to 4.80% average accuracy improvements in federated intra-domain and cross-domain settings.
【7】A Subgoal-driven Framework for Improving Long-Horizon LLM Agents
标题：用于改进长期LLM代理的子目标驱动框架
链接：https://arxiv.org/abs/2603.19685
备注：50 pages, 15 figures
摘要：基于大语言模型（LLM）的代理已经成为数字环境（包括移动界面，操作系统和Web浏览器）的强大自治控制器。例如，Web导航需要处理动态内容和长序列的操作，这使得它特别具有挑战性。现有的基于LLM的代理商主要通过两种方式进行长期规划。在在线执行过程中，当新信息到达时，他们往往会失去跟踪，缺乏通往最终目标的清晰和适应性路径。这个问题在强化学习（RL）微调过程中进一步加剧，稀疏和延迟的奖励使代理难以识别哪些行动导致成功，从而阻止他们在扩展任务中保持连贯的推理。为了应对这些挑战，我们提出了两个贡献。首先，我们引入了一个代理框架，利用专有模型，通过子目标分解在线规划。其次，我们介绍了MiRA（Milestoning Your Reinforcement Learning Enhanced Agent），这是一个强化学习训练框架，使用密集的基于里程碑的奖励信号。实时规划机制改善了专有模型，如Gemini，在WebArena-Lite基准测试中，成功率（SR）绝对增加了约10%。同时，将MiRA应用于开放的Gemma 3 - 12 B模型将其成功率从6.4%提高到43.0%。这一性能超过了专利系统，如GPT-4-Turbo（17.6%）和GPT-4 o（13.9%），以及之前的开放模型最先进的WebRL（38.4%）。总的来说，我们的研究结果表明，结合明确的推理时间规划与基于里程碑的奖励显着提高代理的长期能力，为更强大和通用的自治系统铺平了道路。
摘要：Large language model (LLM)-based agents have emerged as powerful autonomous controllers for digital environments, including mobile interfaces, operating systems, and web browsers. Web navigation, for example, requires handling dynamic content and long sequences of actions, making it particularly challenging. Existing LLM-based agents struggle with long-horizon planning in two main ways. During online execution, they often lose track as new information arrives, lacking a clear and adaptive path toward the final goal. This issue is further exacerbated during reinforcement learning (RL) fine-tuning, where sparse and delayed rewards make it difficult for agents to identify which actions lead to success, preventing them from maintaining coherent reasoning over extended tasks. To address these challenges, we propose two contributions. First, we introduce an agent framework that leverages proprietary models for online planning through subgoal decomposition. Second, we present MiRA (Milestoning your Reinforcement Learning Enhanced Agent), an RL training framework that uses dense, milestone-based reward signals. The real-time planning mechanism improves proprietary models such as Gemini by approximately a 10% absolute increase in success rate (SR) on the WebArena-Lite benchmark. Meanwhile, applying MiRA to the open Gemma3-12B model increases its success rate from 6.4% to 43.0%. This performance surpasses proprietary systems such as GPT-4-Turbo (17.6%) and GPT-4o (13.9%), as well as the previous open-model state of the art, WebRL (38.4%). Overall, our findings demonstrate that combining explicit inference-time planning with milestone-based rewards significantly improves an agent's long-horizon capabilities, paving the way for more robust and general-purpose autonomous systems.
【8】GoAgent: Group-of-Agents Communication Topology Generation for LLM-based Multi-Agent Systems
标题：GoAgent：基于LLM的多Agent系统的Agent组通信布局生成
链接：https://arxiv.org/abs/2603.19677
摘要：基于大语言模型（LLM）的多智能体系统（MAS）在解决复杂任务方面表现出了卓越的能力，但它们的有效性在很大程度上取决于协调智能体交互的底层通信拓扑结构。在这些系统中，成功解决问题往往需要特定任务的小组结构来划分和征服子任务。然而，大多数现有的方法以节点为中心的方式生成通信拓扑，使组结构隐式地从本地连接决策中出现，而不是显式地对其进行建模，这通常会导致次优协调和不必要的通信开销。为了解决这个问题，我们提出了GoAgent（组代理），通信拓扑生成方法，明确对待协作组的MAS建设的原子单位。具体来说，GoAgent首先通过LLM枚举任务相关的候选组，然后自回归选择并连接这些组作为原子单元来构建最终的通信图，共同捕获组内凝聚力和组间协调。为了减轻扩展拓扑中固有的通信冗余和噪声传播，我们进一步引入了一个条件信息瓶颈（CIB）目标，该目标压缩组间通信，保留任务相关信号，同时滤除冗余的历史噪声。在六个基准测试上进行的大量实验证明了GoAgent的最新性能，平均准确率为93.84%，同时减少了约17%的令牌消耗。
摘要：Large language model (LLM)-based multi-agent systems (MAS) have demonstrated exceptional capabilities in solving complex tasks, yet their effectiveness depends heavily on the underlying communication topology that coordinates agent interactions. Within these systems, successful problem-solving often necessitates task-specific group structures to divide and conquer subtasks. However, most existing approaches generate communication topologies in a node-centric manner, leaving group structures to emerge implicitly from local connectivity decisions rather than modeling them explicitly, often leading to suboptimal coordination and unnecessary communication overhead. To address this limitation, we propose GoAgent (Group-of-Agents), a communication topology generation method that explicitly treats collaborative groups as the atomic units of MAS construction. Specifically, GoAgent first enumerates task-relevant candidate groups through an LLM and then autoregressively selects and connects these groups as atomic units to construct the final communication graph, jointly capturing intra-group cohesion and inter-group coordination. To mitigate communication redundancy and noise propagation inherent in expanding topologies, we further introduce a conditional information bottleneck (CIB) objective that compresses inter-group communication, preserving task-relevant signals while filtering out redundant historical noise. Extensive experiments on six benchmarks demonstrate the state-of-the-art performance of GoAgent with 93.84% average accuracy while reducing token consumption by about 17%.
【9】Neural Uncertainty Principle: A Unified View of Adversarial Fragility and LLM Hallucination
标题：神经不确定性原理：对抗脆弱性和LLM幻觉的统一观点
链接：https://arxiv.org/abs/2603.19562
备注：16 pages,3 figures
摘要：大型语言模型中的视觉和幻觉的对抗性脆弱性通常被视为单独的问题，每个问题都用特定于模态的补丁来解决。这项研究首先揭示了它们有一个共同的几何起源：输入和它的损失梯度是共轭可观的不可约的不确定性的限制。在损失诱导状态下形式化神经不确定性原理（NUP），我们发现在近约束状态下，进一步压缩必须伴随着增加的敏感性分散（对抗性脆弱性），而弱梯度耦合使生成受到限制（幻觉）。至关重要的是，这种约束是由输入梯度相关通道，捕获一个专门设计的单反向探头调制。在视觉中，掩蔽高度耦合的组件可以提高鲁棒性，而无需昂贵的对抗训练;在语言中，相同的预填充阶段探测器在生成任何答案令牌之前检测幻觉风险。因此，NUP将两个看似独立的故障分类变成了一个共享的不确定性预算视图，并为可靠性分析提供了一个原则性的镜头。在这个NUP理论的指导下，我们提出了ConjMask（掩蔽高贡献输入分量）和LogitReg（logit-side正则化）来提高鲁棒性，而无需对抗训练，并使用探针作为LLM的无解码风险信号，从而实现幻觉检测和及时选择。因此，NUP提供了一个统一的，实用的框架，用于诊断和减轻感知和生成任务之间的边界异常。
摘要：Adversarial vulnerability in vision and hallucination in large language models are conventionally viewed as separate problems, each addressed with modality-specific patches. This study first reveals that they share a common geometric origin: the input and its loss gradient are conjugate observables subject to an irreducible uncertainty bound. Formalizing a Neural Uncertainty Principle (NUP) under a loss-induced state, we find that in near-bound regimes, further compression must be accompanied by increased sensitivity dispersion (adversarial fragility), while weak prompt-gradient coupling leaves generation under-constrained (hallucination). Crucially, this bound is modulated by an input-gradient correlation channel, captured by a specifically designed single-backward probe. In vision, masking highly coupled components improves robustness without costly adversarial training; in language, the same prefill-stage probe detects hallucination risk before generating any answer tokens. NUP thus turns two seemingly separate failure taxonomies into a shared uncertainty-budget view and provides a principled lens for reliability analysis. Guided by this NUP theory, we propose ConjMask (masking high-contribution input components) and LogitReg (logit-side regularization) to improve robustness without adversarial training, and use the probe as a decoding-free risk signal for LLMs, enabling hallucination detection and prompt selection. NUP thus provides a unified, practical framework for diagnosing and mitigating boundary anomalies across perception and generation tasks.
【10】EvidenceRL: Reinforcing Evidence Consistency for Trustworthy Language Models
标题：EvidenceRL：加强值得信赖的语言模型的证据一致性
链接：https://arxiv.org/abs/2603.19532
摘要：大型语言模型（LLM）是流畅的，但容易产生幻觉，产生看似合理但没有证据支持的答案。这种失败在高风险领域尤其成问题，在这些领域，决策必须以可核实的信息为依据。我们引入\textbf{EvidenceRL}，这是一个强化学习框架，可以在培训期间强制遵守证据。EvidenceRL对候选响应的基础（与检索到的证据和上下文的蕴涵）和正确性（与参考答案的一致性）进行评分，并使用组相对策略优化（GRPO）优化生成器。我们在两个高风险领域进行评估，心脏诊断和法律推理，其中EvidenceRL在不牺牲任务准确性的情况下不断提高证据基础和忠诚度。在心脏诊断方面，Llama-3.2-3B的F1@3从37.0增加到54.5，而接地（$G@3 $）从47.6增加到78.2，幻觉下降了近50%，证据支持的诊断从31.8%增加到61.6%。在法律推理方面，EvidenceRL在Llama-3.1-8B上将忠诚度从32.8%提高到67.6%，表现出跨领域的一致行为变化。我们的代码在https://github.com/Wizaaard/EvidenceRL.git上开源。
摘要：Large Language Models (LLMs) are fluent but prone to hallucinations, producing answers that appear plausible yet are unsupported by available evidence. This failure is especially problematic in high-stakes domains where decisions must be justified by verifiable information. We introduce \textbf{EvidenceRL}, a reinforcement learning framework that enforces evidence adherence during training. EvidenceRL scores candidate responses for grounding (entailment with retrieved evidence and context) and correctness (agreement with reference answers) and optimizes the generator using Group Relative Policy Optimization (GRPO). We evaluate across two high-stakes domains, cardiac diagnosis and legal reasoning, where EvidenceRL consistently improves evidence grounding and faithfulness without sacrificing task accuracy. On cardiac diagnosis, F1@3 increases from 37.0 to 54.5 on Llama-3.2-3B while grounding ($G_{\max}@3$) rises from 47.6 to 78.2; hallucinations drop nearly 5$\times$ and evidence-supported diagnoses increase from 31.8\% to 61.6\%. On legal reasoning, EvidenceRL raises Faithfulness from 32.8\% to 67.6\% on Llama-3.1-8B, demonstrating consistent behavioral change across domains. Our code is open-sourced at https://github.com/Wizaaard/EvidenceRL.git.
【11】Adaptive Layerwise Perturbation: Unifying Off-Policy Corrections for LLM RL
标题：自适应分层扰动：统一LLM RL的政策外纠正
链接：https://arxiv.org/abs/2603.19470
摘要：策略失效、训练-推理不匹配等非策略问题已成为制约LLM强化学习训练稳定性和进一步探索的主要瓶颈。为了提高推理效率，推理和更新策略之间的分布差距增加，导致重尾重要性比。当策略局部尖锐时，会出现重尾比率，这进一步膨胀了尖锐的梯度，并可以将更新推到信任区域之外。为了解决这个问题，我们提出了自适应分层扰动（ALP），通过在更新过程中将小的可学习扰动注入到每层的输入隐藏状态中，该扰动被用作目标中不变推理策略的重要性比的分子。直观地，通过向中间表示添加受控噪声，ALP防止更新的策略过于急剧地偏离推理策略，并扩大策略族以覆盖具有失配噪声的推理策略族。因此，扁平化分布可以自然地缩小更新和推理策略的差距，并减少重要性比率的尾部，从而保持训练稳定性。这是进一步验证经验。在单轮数学和多轮工具集成推理任务上的实验表明，ALP不仅提高了最终性能，而且在迭代训练过程中避免了重要性比尾部和KL尖峰的爆炸，同时促进了探索。烧蚀表明，所有层的代表性水平的扰动是最有效的，大大优于部分层和logits只有变种。
摘要：Off-policy problems such as policy staleness and training-inference mismatch, has become a major bottleneck for training stability and further exploration for LLM RL. To enhance inference efficiency, the distribution gap between the inference and updated policy grows, leading to heavy-tailed importance ratios. Heavy-tailed ratios arise when the policy is locally sharp, which further inflates sharp gradients and can push updates outside the trust region. To address this, we propose Adaptive Layerwise Perturbation(ALP) by injecting small learnable perturbations into input hidden states of each layer during updates, which is used as the numerator of the importance ratio against the unchanged inference policy in the objective. Intuitively, by adding controlled noise to intermediate representations, ALP prevents the updated policy from deviating too sharply from the inference policy, and enlarges the policy family to cover the inference policy family with mismatch noises. Hence, the flattened distribution can naturally tighten the updated and inference policy gap and reduce the tail of importance ratios, thus maintaining training stability. This is further validated empirically. Experiments on single-turn math and multi-turn tool-integrated reasoning tasks show that ALP not only improves final performance, but also avoid blow up of importance ratio tail and KL spikes during iterative training, along with boosted exploration. Ablations show that representation-level perturbations across all layers are most effective, substantially outperforming partial-layer and logits-only variants.
【12】GeoLAN: Geometric Learning of Latent Explanatory Directions in Large Language Models
标题：地理LAN：大型语言模型中潜在解释方向的几何学习
链接：https://arxiv.org/abs/2603.19460
摘要：大型语言模型（LLM）表现出强大的性能，但它们往往缺乏透明度。我们引入GeoLAN，这是一个训练框架，它将令牌表示视为几何轨迹，并应用受最近与Kakeya猜想相关的发展启发的粘性条件。我们已经开发了两个可微正则化器，Katz-Tao Convex Wolff（KT-CW）和Katz-Tao Attention（KT-Attn），它们促进各向同性并鼓励多样化的注意力。我们对Gemma-3（1B，4 B，12 B）和Llama-3-8B的实验表明，GeoLAN经常保持任务的准确性，同时提高几何度量并减少某些公平性偏差。这些优势在中型车型中最为显著。我们的研究结果揭示了几何精度和性能之间的尺度依赖性权衡，这表明几何感知训练是一种很有前途的方法，以提高机械可解释性。
摘要：Large language models (LLMs) demonstrate strong performance, but they often lack transparency. We introduce GeoLAN, a training framework that treats token representations as geometric trajectories and applies stickiness conditions inspired by recent developments related to the Kakeya Conjecture. We have developed two differentiable regularizers, Katz-Tao Convex Wolff (KT-CW) and Katz-Tao Attention (KT-Attn), that promote isotropy and encourage diverse attention. Our experiments with Gemma-3 (1B, 4B, 12B) and Llama-3-8B show that GeoLAN frequently maintains task accuracy while improving geometric metrics and reducing certain fairness biases. These benefits are most significant in mid-sized models. Our findings reveal scale-dependent trade-offs between geometric precision and performance, suggesting that geometry-aware training is a promising approach to enhance mechanistic interpretability.
【13】Vocabulary shapes cross-lingual variation of word-order learnability in language models
标题：词汇塑造了语言模型中语序可学习性的跨语言差异
链接：https://arxiv.org/abs/2603.19427
备注：Submitted to ACL 2026. 17 pages, 11 figures
摘要：为什么有些语言如捷克语允许自由词序，而有些语言如英语则不允许？我们解决这个问题的预训练Transformer语言模型的合成自然语言的词序变体的频谱。我们观察到，更大的词序不规则性不断提高模型的可学习性，表明降低。然而，句子颠倒对可学性的影响很小。自由的粗略区别（例如，捷克语和芬兰语）和固定词序语言（例如，英语和法语）并不能解释跨语言的变异。相反，单词和子单词词汇的结构强烈地预测了模型的语义。总体而言，词汇结构成为跨语言计算语序可学习性的关键驱动力。
摘要：Why do some languages like Czech permit free word order, while others like English do not? We address this question by pretraining transformer language models on a spectrum of synthetic word-order variants of natural languages. We observe that greater word-order irregularity consistently raises model surprisal, indicating reduced learnability. Sentence reversal, however, affects learnability only weakly. A coarse distinction of free- (e.g., Czech and Finnish) and fixed-word-order languages (e.g., English and French) does not explain cross-lingual variation. Instead, the structure of the word and subword vocabulary strongly predicts the model surprisal. Overall, vocabulary structure emerges as a key driver of computational word-order learnability across languages.
【14】The Autonomy Tax: Defense Training Breaks LLM Agents
标题：自主税：国防训练打破了法学硕士代理人的困境
链接：https://arxiv.org/abs/2603.19423
摘要：大型语言模型（LLM）代理越来越依赖外部工具（文件操作，API调用，数据库事务）来自主完成复杂的多步任务。从业者部署经过防御训练的模型，以防止通过恶意观察或检索内容操纵代理行为的即时注入攻击。我们揭示了一个基本的\textbf{能力对齐悖论}：旨在提高安全性的防御训练系统地破坏代理能力，而未能防止复杂的攻击。在97个代理任务和1,000个对抗性提示中，针对不设防基线评估防御模型，我们发现了多步代理特有的三个系统性偏差。\textbf{Agent incompetence bias}表现为即时工具执行故障，模型在观察任何外部内容之前拒绝或生成对良性任务的无效操作。\textbf{Cascade amplification bias}会导致早期失败通过重试循环传播，使防御模型在99%的任务上超时，而基线为13%。\textbf{Trigger bias}会导致自相矛盾的安全性下降，其中防御模型的性能比未防御的基线更差，而直接攻击则以高速率绕过防御。根本原因分析表明，这些偏见源于捷径学习：模型过度拟合表面攻击模式，而不是语义威胁理解，这一点可以从不同攻击类别的防御有效性差异中得到证明。我们的研究结果表明，目前的防御模式优化了单回合拒绝基准，同时使多步代理从根本上不可靠，需要新的方法，在对抗条件下保持工具执行能力。
摘要：Large language model (LLM) agents increasingly rely on external tools (file operations, API calls, database transactions) to autonomously complete complex multi-step tasks. Practitioners deploy defense-trained models to protect against prompt injection attacks that manipulate agent behavior through malicious observations or retrieved content. We reveal a fundamental \textbf{capability-alignment paradox}: defense training designed to improve safety systematically destroys agent competence while failing to prevent sophisticated attacks. Evaluating defended models against undefended baselines across 97 agent tasks and 1,000 adversarial prompts, we uncover three systematic biases unique to multi-step agents. \textbf{Agent incompetence bias} manifests as immediate tool execution breakdown, with models refusing or generating invalid actions on benign tasks before observing any external content. \textbf{Cascade amplification bias} causes early failures to propagate through retry loops, pushing defended models to timeout on 99\% of tasks compared to 13\% for baselines. \textbf{Trigger bias} leads to paradoxical security degradation where defended models perform worse than undefended baselines while straightforward attacks bypass defenses at high rates. Root cause analysis reveals these biases stem from shortcut learning: models overfit to surface attack patterns rather than semantic threat understanding, evidenced by extreme variance in defense effectiveness across attack categories. Our findings demonstrate that current defense paradigms optimize for single-turn refusal benchmarks while rendering multi-step agents fundamentally unreliable, necessitating new approaches that preserve tool execution competence under adversarial conditions.
【15】Automated Membership Inference Attacks: Discovering MIA Signal Computations using LLM Agents
标题：自动成员资格推理攻击：使用LLM代理发现MIA信号计算
链接：https://arxiv.org/abs/2603.19375
摘要：成员推理攻击（MIA）使攻击者能够确定特定数据点是否是模型训练数据集的一部分，已成为理解、评估和量化与机器学习系统相关的潜在信息泄漏的重要框架。设计有效的MIA是一项具有挑战性的任务，通常需要对模型行为进行大量的手动探索，以识别潜在的漏洞。在本文中，我们介绍AutoMIA -一种新的框架，利用大语言模型（LLM）代理自动化的设计和实现新的MIA信号计算。通过利用LLM代理，我们可以系统地探索潜在攻击策略的广阔空间，从而发现新的策略。我们的实验表明，AutoMIA可以成功地发现专门针对用户配置的目标模型和数据集定制的新MIA，从而使绝对AUC比现有MIA提高高达0.18。这项工作首次证明了LLM代理可以作为设计和实现具有SOTA性能的MIA的有效且可扩展的范例，为未来的探索开辟了新的途径。
摘要：Membership inference attacks (MIAs), which enable adversaries to determine whether specific data points were part of a model's training dataset, have emerged as an important framework to understand, assess, and quantify the potential information leakage associated with machine learning systems. Designing effective MIAs is a challenging task that usually requires extensive manual exploration of model behaviors to identify potential vulnerabilities. In this paper, we introduce AutoMIA -- a novel framework that leverages large language model (LLM) agents to automate the design and implementation of new MIA signal computations. By utilizing LLM agents, we can systematically explore a vast space of potential attack strategies, enabling the discovery of novel strategies. Our experiments demonstrate AutoMIA can successfully discover new MIAs that are specifically tailored to user-configured target model and dataset, resulting in improvements of up to 0.18 in absolute AUC over existing MIAs. This work provides the first demonstration that LLM agents can serve as an effective and scalable paradigm for designing and implementing MIAs with SOTA performance, opening up new avenues for future exploration.
【16】Anatomical Heterogeneity in Transformer Language Models
标题：Transformer语言模型中的解剖学异类
链接：https://arxiv.org/abs/2603.19348
备注：11 pages, 10 tables. Independent research. Code available at https://github.com/tomaszwi66
摘要：电流Transformer语言模型在所有层上使用统一的计算预算进行训练，隐含地假设层同质性。我们通过对SmolLM 2 - 135 M（30层，135 M参数因果语言模型）的实证分析来挑战这一假设，使用五个诊断指标：权重可预测性（R2），消融退化，恢复速度，权重操纵鲁棒性和结构分析。我们发现了深刻的解剖异质性：（1）层重量遵循强的数学规律性（R2 = 0.91）与普遍的振荡三角洲模式（相关性~= -0.50），但预测的重量导致灾难性的失败，由于非线性误差积累。(2)层的重要性跨越10^7的范围，从关键核心（L 8 -11，高达+63，419%PPL降解）到反层（L14，L17），其去除可提高性能。(3)恢复速度与层的重要性相关，表明不同的训练要求。(4)在五种测试的操作策略中，只有权重缩放（alpha = 0.9）保留了模型质量。(5)增长型Transformer培训，按层重要性分配预算，实现约54%的成本降低。概念验证实验证实了这一点：在相同的参数计数下，验证损失比均匀训练低4.7倍，同时速度快13%。
摘要：Current transformer language models are trained with uniform computational budgets across all layers, implicitly assuming layer homogeneity. We challenge this assumption through empirical analysis of SmolLM2-135M, a 30-layer, 135M-parameter causal language model, using five diagnostic metrics: weight predictability (R2), ablation degradation, recovery speed, weight manipulation robustness, and structural analysis. We find profound anatomical heterogeneity: (1) Layer weights follow strong mathematical regularity (R2 = 0.91) with a universal oscillatory delta pattern (correlation ~= -0.50), yet predicted weights cause catastrophic failure due to nonlinear error accumulation. (2) Layer importance spans a 10^7 range, from a critical core (L8-11, up to +63,419% PPL degradation) to anti-layers (L14, L17) whose removal improves performance. (3) Recovery speed correlates with layer importance, indicating differential training requirements. (4) Only weight scaling (alpha = 0.9) preserves model quality among five tested manipulation strategies. (5) Growth Transformer Training, allocating budget by layer importance, achieves ~54% cost reduction. A proof-of-concept experiment confirms this: 4.7x lower validation loss than uniform training at identical parameter count, while being 13% faster.
【17】MemReward: Graph-Based Experience Memory for LLM Reward Prediction with Limited Labels
标题：MemReward：基于图形的经验记忆，用于LLM奖励预测，标签有限
链接：https://arxiv.org/abs/2603.19310
摘要：通过强化学习训练用于复杂推理的大型语言模型（LLM）需要奖励标签来指定生成的卷展是否正确。然而，大规模获得奖励标签通常需要昂贵的人工标签或耗时的验证程序;例如，评估数学证明需要专家评审，而开放式问题回答缺乏明确的基础事实。当奖励标签有限时，强化学习微调的有效性受到奖励标签稀缺性的限制。我们引入MemReward，一个基于图的经验记忆框架：初始LLM策略为每个查询生成卷展，每个查询包括一个思考过程和一个最终答案，这些卷展存储为经验记忆。搜索、思考过程和答案形成了具有相似性和结构边的异构图中的节点;在标记节点上训练的GNN在在线优化期间将奖励传播给未标记的推出。在Qwen2.5-3B和1.5B上进行的数学、问题回答和代码生成实验表明，MemReward仅使用20%的标签，在3B和1.5B上的性能分别达到了Oracle的97.3%和96.6%，在域外任务上超过了Oracle。性能随标签预算的变化而平稳扩展，在70%标签的情况下达到Oracle的99.4%。
摘要：Training large language models (LLMs) for complex reasoning via reinforcement learning requires reward labels that specify whether the generated rollouts are correct. However, obtaining reward labels at scale often requires expensive human labeling or time-consuming verification procedures; for instance, evaluating mathematical proofs demands expert review, while open-ended question answering lacks definitive ground truth. When reward labels are limited, the effectiveness of reinforcement learning fine-tuning is constrained by the scarcity of reward labels. We introduce MemReward, a graph-based experience memory framework: an initial LLM policy generates rollouts for each query, each comprising a thinking process and a final answer, and these rollouts are stored as experience memory. Queries, thinking processes, and answers form nodes in a heterogeneous graph with similarity and structural edges; a GNN trained on labeled nodes propagates rewards to unlabeled rollouts during online optimization. Experiments on Qwen2.5-3B and 1.5B across mathematics, question answering, and code generation demonstrate that MemReward, with only 20% labels, achieves 97.3% of Oracle performance on 3B and 96.6% on 1.5B, surpassing Oracle on out-of-domain tasks. Performance scales smoothly with label budget, reaching 99.4% of Oracle at 70% labels.
【18】CLaRE-ty Amid Chaos: Quantifying Representational Entanglement to Predict Ripple Effects in LLM Editing
标题：混乱中的CLaRE-ty：量化代表性纠缠以预测LLM编辑中的涟漪效应
链接：https://arxiv.org/abs/2603.19297
摘要：随着时间的推移，大型语言模型（LLM）的静态知识表示不可避免地变得过时或不正确。虽然模型编辑技术通过修改模型的事实关联提供了一个有前途的解决方案，但它们通常会产生不可预测的涟漪效应，这是一种无意的行为变化，甚至会传播到隐藏的空间。在这项工作中，我们介绍了CLaRE，一个轻量级的代表级技术，以确定这些涟漪效应可能发生。与以前的基于梯度的方法不同，CLaRE使用来自单个中间层的前向激活来量化事实之间的纠缠，避免了昂贵的后向传递。为了进行系统的研究，我们准备和分析了从三个现有数据集中提取的11，427个事实的语料库。使用CLaRE，我们计算这个语料库的多个模型的大规模纠缠图，捕捉本地编辑如何通过代表性空间传播。这些图表为模型编辑、审计跟踪、高效的红团队和可扩展的编辑后评估提供了更强大的保留集。与基线相比，CLaRE在Spearman相关性方面实现了平均62.2%的改进，同时速度提高了2.74倍，使用的峰值GPU内存减少了2.85倍。此外，CLaRE只需要基线所需的一小部分存储来计算和保存事实表示。我们的纠缠图和语料库可以在https://anonymous.4open.science/r/CLaRE-488E上找到。
摘要：The static knowledge representations of large language models (LLMs) inevitably become outdated or incorrect over time. While model-editing techniques offer a promising solution by modifying a model's factual associations, they often produce unpredictable ripple effects, which are unintended behavioral changes that propagate even to the hidden space. In this work, we introduce CLaRE, a lightweight representation-level technique to identify where these ripple effects may occur. Unlike prior gradient-based methods, CLaRE quantifies entanglement between facts using forward activations from a single intermediate layer, avoiding costly backward passes. To enable systematic study, we prepare and analyse a corpus of 11,427 facts drawn from three existing datasets. Using CLaRE, we compute large-scale entanglement graphs of this corpus for multiple models, capturing how local edits propagate through representational space. These graphs enable stronger preservation sets for model editing, audit trails, efficient red-teaming, and scalable post-edit evaluation. In comparison to baselines, CLaRE achieves an average of 62.2% improvement in Spearman correlation with ripple effects while being $2.74\times$ faster, and using $2.85\times$ less peak GPU memory. Besides, CLaRE requires only a fraction of the storage needed by the baselines to compute and preserve fact representations. Our entanglement graphs and corpus are available at https://anonymous.4open.science/r/CLaRE-488E.
【19】TTQ: Activation-Aware Test-Time Quantization to Accelerate LLM Inference On The Fly
标题：TTQ：激活感知测试时间量化，以加速LLM动态推理
链接：https://arxiv.org/abs/2603.19296
备注：25 pages
摘要：为了解决大型基础模型的巨大计算需求，已经引入了无需再训练的激活感知压缩技术。然而，由于这些方法高度依赖于校准数据，因此对于看不见的下游任务可能会出现域偏移问题。我们提出了一个测试时间量化（TTQ）框架，压缩大型模型的飞行在推理时间来解决这个问题。通过有效的在线校准，即时激活感知量化可以适应每个提示，而不管下游任务如何，同时实现推理加速。几个实验表明，TTQ可以提高量化性能超过国家的最先进的基线。
摘要：To tackle the huge computational demand of large foundation models, activation-aware compression techniques without retraining have been introduced. However, since these methods highly rely on calibration data, domain shift issues may arise for unseen downstream tasks. We propose a test-time quantization (TTQ) framework which compresses large models on the fly at inference time to resolve this issue. With an efficient online calibration, instant activation-aware quantization can adapt every prompt regardless of the downstream tasks, yet achieving inference speedup. Several experiments demonstrate that TTQ can improve the quantization performance over state-of-the-art baselines.
【20】Maximizing mutual information between user-contexts and responses improve LLM personalization with no additional data
标题：最大化用户上下文和响应之间的相互信息，无需额外数据即可改善LLM个性化
链接：https://arxiv.org/abs/2603.19294
摘要：虽然后训练已经成功地改善了各种领域的大型语言模型（LLM），但这些收益严重依赖于人类标记的数据或外部验证器。现有的数据已经被利用，新的高质量数据的收集成本很高。更根本的是，真正的情报远远超出了容易验证的任务。因此，我们需要自我改进框架，允许模型在没有外部监督的情况下改进。我们提出了 * 互信息偏好优化（MIPO）*，一种对比数据增强方法，通过在正确提示上生成积极响应条件来构建偏好对，并通过在随机，不相关提示上生成消极响应条件来构建偏好对。我们表明，使用直接偏好优化（DPO），从这个配对的数据学习最大化逐点条件互信息（MI）（在基础LLM下）之间的提示和模型响应。不同大小的Llama和Qwen-Instruct模型的实证结果表明，当用于最大化用户上下文和响应之间的MI时，MIPO提供了一种有效的个性化技术，与强基线相比，使用真实用户数据集在个性化任务上实现了3-40%的改进。令人惊讶的是，MIPO还可以用于提高数学和多项选择题的成绩，在没有任何额外数据或人工监督的情况下，可以提高1-18% **。这些结果为自我改进提供了一个有希望的方向。
摘要 ：While post-training has successfully improved large language models (LLMs) across a variety of domains, these gains heavily rely on human-labeled data or external verifiers. Existing data has already been exploited, and new high-quality data is expensive to collect. More fundamentally, true intelligence goes far beyond tasks that are easily verifiable. Therefore, we need self-improvement frameworks that allow models to improve without external oversight. We propose *Mutual Information Preference Optimization (MIPO)*, a contrastive data augmentation method that constructs preference pairs by generating a positive response conditioning on the correct prompt, and a negative response by conditioning on a random, unrelated prompt. We show that using Direct Preference Optimization (DPO) to learn from this paired data maximizes pointwise conditional mutual information (MI) (under the base LLM) between prompts and model responses. Empirical results with various-sized Llama- and Qwen-Instruct models show that when used to maximize MI between user context and response, MIPO provides an effective personalization technique, achieving 3-40% improvements on personalization tasks using real-user datasets compared to strong baselines. Surprisingly, MIPO can also be applied to improve performance on math and multiple-choice problems, yielding 1-18% **without any additional data or human supervision**. These results suggest a promising direction for self-improvement.
【21】Probing to Refine: Reinforcement Distillation of LLMs via Explanatory Inversion
标题：探索精炼：通过解释倒置强化蒸馏LLM
链接：https://arxiv.org/abs/2603.19266
备注：Accepted to ICLR 2026
摘要：将强大的推理能力从大型语言模型（LLM）中提取到较小的，计算效率高的学生模型仍然是一个尚未解决的挑战。尽管最近取得了进展，但提取的模型经常遭受肤浅的模式记忆和低于标准的概括。为了克服这些限制，我们引入了一个新的蒸馏框架，超越简单的模仿灌输更深层次的概念理解。我们的框架具有两个关键创新。\underline {\texit {First}}，为了解决模式记忆问题，解释性倒置（EI）生成有针对性的"解释性问题"，迫使学生阐明答案背后的逻辑，而不仅仅是记住它。\underline {\texit {Second}}，为了提高概括能力，解释性GRPO（\texttt {EXGRPO}）使用了一种强化学习算法，具有一种新颖的对话结构效用奖励，它明确奖励学生在这些探测器中保持连贯的推理过程。对12个数据集的广泛评价显示出显著的改进。使用Gemma-7b作为学生模型，我们的方法比zero-shot性能平均提高了20.39\ %，比最先进的蒸馏基线提高了6.02\ %。此外，用我们的方法提取的模型显示出显著的训练效率（例如，超越了使用\textbf {10 - 25\%}训练数据进行的普通微调），并对分布外任务具有很强的泛化能力。实现在www.example.com上发布。
摘要：Distilling robust reasoning capabilities from large language models (LLMs) into smaller, computationally efficient student models remains an unresolved challenge. Despite recent advances, distilled models frequently suffer from superficial pattern memorization and subpar generalization. To overcome these limitations, we introduce a novel distillation framework that moves beyond simple mimicry to instill a deeper conceptual understanding. Our framework features two key innovations. \underline{\textit{First}}, to address pattern memorization, Explanatory Inversion (EI) generates targeted ``explanatory probes'' that compel the student to articulate the underlying logic behind an answer, rather than just memorizing it. \underline{\textit{Second}}, to improve generalization, Explanatory GRPO (\texttt{EXGRPO}) uses a reinforcement learning algorithm with a novel Dialogue Structure Utility Bonus, which explicitly rewards the student for maintaining a coherent reasoning process across these probes. Extensive evaluations on 12 datasets demonstrate significant improvements. Using Gemma-7b as the student model, our method yields an average \textbf{20.39\%} increase over zero-shot performance and a \textbf{6.02\%} improvement over the state-of-the-art distillation baselines. Moreover, models distilled with our method show remarkable training efficiency (e.g., surpassing vanilla fine-tuning with \textbf{10-25\%} training data) and strong generalization to out-of-distribution tasks. Implementation is released at https://github.com/Zhen-Tan-dmml/ExGRPO.git.
【22】Significance-Gain Pair Encoding for LLMs: A Statistical Alternative to Frequency-Based Subword Merging
标题：LLM的有效性-收益对编码：基于频率的子字合并的统计替代方案
链接：https://arxiv.org/abs/2603.19261
备注：8 pages, 1 figures
摘要：子字标记化是现代语言模型（包括大型语言模型（LLM））的关键设计选择，字节和字符级BPE作为广泛使用的基线。标准BPE通过原始对频率来选择合并，这有利于压缩，但可能会将真正的邻接内聚与由于高边缘计数而频繁出现的对混为一谈。本文介绍了显着增益BPE，一个下拉式的替代合并标准，通过独立空模型下的z统计量来衡量凝聚力，并将其与显式的压缩感知增益项相结合。显著性增益BPE使用小型因果Transformer语言模型在WikiText-103（原始）字符切片上进行评估，报告令牌相关的困惑和令牌化器不变的每个字符的度量位（BPC）。在一个代表性的操作点，显著性增益BPE分别将验证和测试困惑降低了13%和12%，并将验证和测试BPC提高了约0.9%至1.0%。词汇表大小扫描进一步显示在最接近的压缩比较中较低的BPC，这表明统计接地合并选择可以在一系列压缩机制中提高每单位原始文本的预测效率。
摘要：Subword tokenization is a key design choice for modern language models, including large language models (LLMs), with byte- and character-level BPE serving as a widely used baseline. Standard BPE selects merges by raw pair frequency, which favors compression but can conflate true adjacency cohesion with pairs that are frequent due to high marginal counts. This paper introduces Significance-Gain BPE, a drop-in alternative merge criterion that measures cohesion via a z-statistic under an independence null model and combines it with an explicit compression-aware gain term. Significance-Gain BPE is evaluated on WikiText-103 (raw) character slices using a small causal Transformer language model, reporting both token-dependent perplexity and the tokenizer-invariant metric bits per character (BPC). At a representative operating point, Significance-Gain BPE reduces validation and test perplexity by 13% and 12%, respectively, and improves validation and test BPC by about 0.9 to 1.0%. A vocabulary-size sweep further shows lower BPC in most closest-compression comparisons, suggesting that statistically grounded merge selection can improve predictive efficiency per unit of raw text across a range of compression regimes.
【23】Reinforcement-guided generative protein language models enable de novo design of highly diverse AAV capsids
标题：增强引导的生成蛋白质语言模型能够重新设计高度多样化的AAC外壳
链接：https://arxiv.org/abs/2603.19473
摘要：腺相关病毒（AAV）载体是基因治疗中广泛使用的递送平台，设计改进的衣壳是扩大其治疗潜力的关键。在AAV生物工程中的一个中心挑战，更广泛地说，在蛋白质设计中，是相对于可行的实验筛选规模的巨大序列设计空间。机器引导的生成方法提供了一种强大的手段来导航这一景观，并提出满足功能约束的新蛋白质序列。在这里，我们开发了一个基于蛋白质语言模型和强化学习的生成设计框架，以生成高度新颖但功能上合理的AAV衣壳。预训练的模型在实验验证的衣壳序列上进行微调，以学习与活力相关的模式。然后使用强化学习来指导序列生成，并使用奖励函数联合促进预测的可行性和序列新颖性，从而实现对训练数据中所代表区域的探索。比较分析表明，单独进行微调会产生具有高预测生存力的序列，但仍然偏向于训练分布，而强化学习引导的生成会到达序列空间的更远区域，同时保持高预测生存力。最后，我们提出了一种候选选择策略，该策略整合了预测的生存能力，序列新颖性和生物物理特性，以优先考虑下游评估的变体。这项工作建立了一个框架，为蛋白质序列空间的生成探索，并推进了生成蛋白质语言模型的应用AAV生物工程。
摘要：Adeno-associated viral (AAV) vectors are widely used delivery platforms in gene therapy, and the design of improved capsids is key to expanding their therapeutic potential. A central challenge in AAV bioengineering, as in protein design more broadly, is the vast sequence design space relative to the scale of feasible experimental screening. Machine-guided generative approaches provide a powerful means of navigating this landscape and proposing novel protein sequences that satisfy functional constraints. Here, we develop a generative design framework based on protein language models and reinforcement learning to generate highly novel yet functionally plausible AAV capsids. A pretrained model was fine-tuned on experimentally validated capsid sequences to learn patterns associated with viability. Reinforcement learning was then used to guide sequence generation, with a reward function that jointly promoted predicted viability and sequence novelty, thereby enabling exploration beyond regions represented in the training data. Comparative analyses showed that fine-tuning alone produces sequences with high predicted viability but remains biased toward the training distribution, whereas reinforcement learining-guided generation reaches more distant regions of sequence space while maintaining high predicted viability. Finally, we propose a candidate selection strategy that integrates predicted viability, sequence novelty, and biophysical properties to prioritize variants for downstream evaluation. This work establishes a framework for the generative exploration of protein sequence space and advances the application of generative protein language models to AAV bioengineering.
Graph相关(图学习|图神经网络|图优化等)(8篇)
【1】Graph2TS: Structure-Controlled Time Series Generation via Quantile-Graph VAEs
标题：Graph2TS：通过分位图VAE生成结构控制的时间序列
链接：https://arxiv.org/abs/2603.19970
摘要：虽然最近的生成模型可以产生具有接近边缘分布的时间序列，但它们通常面临着保持全局时间结构和建模随机局部变化之间的根本紧张关系，特别是对于具有弱或不规则周期性的高度波动信号。在这种情况下，直接分布匹配可能会放大噪声或抑制有意义的时间模式。在这项工作中，我们提出了一个结构残差的角度来看时间序列的生成，查看时间数据的结构骨干和随机残差动态的组合，从而激励分离的全球组织样本级的变异性。基于这一见解，我们使用基于分位数的转换图来表示时间序列结构，该图紧凑地捕获了全局分布和时间依赖性。在此表示的基础上，我们提出了Graph2TS，一个分位数图条件变分自编码器，它可以从结构图到时间序列进行跨模态生成。通过对结构而不是标签或元数据进行调节，该模型保留了全局时间组织，同时实现了受控的随机变化。在不同数据集上的实验，包括太阳黑子，电力负荷，ECG和EEG信号，与基于扩散和GAN的基线相比，证明了改进的分布保真度，时间对齐和代表性，突出了结构控制和交叉模态生成作为时间序列建模的一个有前途的方向。
摘要：Although recent generative models can produce time series with close marginal distributions, they often face a fundamental tension between preserving global temporal structure and modeling stochastic local variations, particularly for highly volatile signals with weak or irregular periodicity. Direct distribution matching in such settings can amplify noise or suppress meaningful temporal patterns. In this work, we propose a structure-residual perspective on time-series generation, viewing temporal data as the combination of a structural backbone and stochastic residual dynamics, thereby motivating the separation of global organization from sample-level variability. Based on this insight, we represent time-series structure using a quantile-based transition graph that compactly captures global distributional and temporal dependencies. Building on this representation, we propose Graph2TS, a quantile-graph conditioned variational autoencoder that performs cross-modal generation from structural graphs to time series. By conditioning generation on structure rather than labels or metadata, the model preserves global temporal organization while enabling controlled stochastic variation. Experiments on diverse datasets, including sunspot, electricity load, ECG, and EEG signals, demonstrate improved distributional fidelity, temporal alignment, and representativeness compared to diffusion- and GAN-based baselines, highlighting structure-controlled and cross-modal generation as a promising direction for time-series modeling.
【2】Integrating Meta-Features with Knowledge Graph Embeddings for Meta-Learning
标题：将元特征与知识图嵌入集成以实现元学习
链接：https://arxiv.org/abs/2603.19888
摘要：网络上大量的机器学习记录为元学习提供了一个重要的机会，利用过去的实验来提高性能。两个关键的元学习任务是管道性能估计（PPE），它预测目标数据集上的管道性能，以及基于数据集性能的相似性估计（DPSE），它识别具有相似性能模式的数据集。现有方法主要依赖于数据集元特征（例如，实例数、类熵等）来用数字表示数据集并近似这些元学习任务。然而，这些方法往往忽略了过去的实验结果和管道元数据的财富。这限制了它们捕获数据集-管道交互的能力，这些交互揭示了性能相似性模式。在这项工作中，我们提出了KGmetaSP，知识图嵌入的方法，利用现有的实验数据来捕捉这些相互作用，并提高PPE和DPSE。我们在统一的知识图（KG）中表示数据集和管道，并导出支持PPE的管道不可知元模型和DPSE的基于距离的检索的嵌入。为了验证我们的方法，我们构建了一个包含144，177个OpenML实验的大规模基准测试，从而实现了丰富的跨数据集评估。KGmetaSP使用单个管道无关元模型实现精确的PPE，并在基线上改进DPSE。拟议的KGmetaSP、KG和基准测试已经发布，为元学习建立了一个新的参考点，并展示了如何将开放实验数据整合到一个统一的KG中，从而推动该领域的发展。
摘要：The vast collection of machine learning records available on the web presents a significant opportunity for meta-learning, where past experiments are leveraged to improve performance. Two crucial meta-learning tasks are pipeline performance estimation (PPE), which predicts pipeline performance on target datasets, and dataset performance-based similarity estimation (DPSE), which identifies datasets with similar performance patterns. Existing approaches primarily rely on dataset meta-features (e.g., number of instances, class entropy, etc.) to represent datasets numerically and approximate these meta-learning tasks. However, these approaches often overlook the wealth of past experimental results and pipeline metadata available. This limits their ability to capture dataset - pipeline interactions that reveal performance similarity patterns. In this work, we propose KGmetaSP, a knowledge-graph-embeddings approach that leverages existing experiment data to capture these interactions and improve both PPE and DPSE. We represent datasets and pipelines within a unified knowledge graph (KG) and derive embeddings that support pipeline-agnostic meta-models for PPE and distance-based retrieval for DPSE. To validate our approach, we construct a large-scale benchmark comprising 144,177 OpenML experiments, enabling a rich cross-dataset evaluation. KGmetaSP enables accurate PPE using a single pipeline-agnostic meta-model and improves DPSE over baselines. The proposed KGmetaSP, KG, and benchmark are released, establishing a new reference point for meta-learning and demonstrating how consolidating open experiment data into a unified KG advances the field.
【3】GDEGAN: Gaussian Dynamic Equivariant Graph Attention Network for Ligand Binding Site Prediction
标题：GDEGAN：用于配体结合位点预测的高斯动态等变图注意力网络
链接：https://arxiv.org/abs/2603.19817
摘要：准确预测配体与蛋白质的结合位点是基于结构的计算药物发现的关键步骤。最近，等变图神经网络（GNNs）已经成为结合位点识别方法的强大范例，这是由于通过蛋白质数据库和AlphaFold预测大规模提供蛋白质的3D结构。最先进的等变GNN方法实现了点积注意力，忽略了相邻残基的化学和几何性质的变化。为了捕捉这种变化，我们提出了GDEGAN（高斯动态等变图注意力网络），它用识别结合位点的自适应内核取代了点积注意力。所提出的注意力机制捕获的变化，在相邻的残基使用其特性的局部特征分布的统计。我们的机制在每一层动态计算邻域统计，使用局部方差作为自适应带宽参数，具有可学习的头部温度，使每个蛋白质区域能够确定其自身的上下文特定的重要性。GDEGAN优于现有方法，在COACH 420、HOLO 4k和PDBBind 2020数据集上，DCC和DCA成功率分别提高了37-66%和7-19%。这些进展直接应用于通过鉴定用于治疗靶点鉴定的潜在结合位点来加速蛋白质-配体对接。
摘要：Accurate prediction of binding sites of a given protein, to which ligands can bind, is a critical step in structure-based computational drug discovery. Recently, Equivariant Graph Neural Networks (GNNs) have emerged as a powerful paradigm for binding site identification methods due to the large-scale availability of 3D structures of proteins via protein databases and AlphaFold predictions. The state-of-the-art equivariant GNN methods implement dot product attention, disregarding the variation in the chemical and geometric properties of the neighboring residues. To capture this variation, we propose GDEGAN (Gaussian Dynamic Equivariant Graph Attention Network), which replaces dot-product attention with adaptive kernels that recognize binding sites. The proposed attention mechanism captures variation in neighboring residues using statistics of their characteristic local feature distributions. Our mechanism dynamically computes neighborhood statistics at each layer, using local variance as an adaptive bandwidth parameter with learnable per-head temperatures, enabling each protein region to determine its own context-specific importance. GDEGAN outperforms existing methods with relative improvements of 37-66% in DCC and 7-19% DCA success rates across COACH420, HOLO4k, and PDBBind2020 datasets. These advances have direct application in accelerating protein-ligand docking by identifying potential binding sites for therapeutic target identification.
【4】Stochastic Sequential Decision Making over Expanding Networks with Graph Filtering
标题：图过滤扩展网络的随机顺序决策
链接
：https://arxiv.org/abs/2603.19501
摘要：图过滤器利用拓扑信息来处理网络数据，现有的方法主要研究固定图，忽略了图往往会随着节点以未知模式不断附着而扩展。后者需要开发基于过滤器的决策范式，将进化和不确定性考虑在内。现有的方法依赖于预先设计的过滤器或在线学习，仅限于考虑过去或现在的信息的近视视图。考虑到未来的影响，我们提出了一个随机顺序决策框架过滤网络数据的政策，适应过滤扩展图。通过将过滤器移位表示为代理，我们将过滤器建模为多代理系统，并在多代理强化学习之后训练策略。这说明了长期回报，并通过顺序决策捕捉扩张动态。此外，我们开发了一个上下文感知的图神经网络参数化的政策，调整过滤器参数的基础上的图和代理的信息。在从冷启动推荐到COVID预测的合成和真实数据集上进行的实验突出了使用顺序决策视角优于批量和在线过滤替代方案的好处。
摘要：Graph filters leverage topological information to process networked data with existing methods mainly studying fixed graphs, ignoring that graphs often expand as nodes continually attach with an unknown pattern. The latter requires developing filter-based decision-making paradigms that take evolution and uncertainty into account. Existing approaches rely on either pre-designed filters or online learning, limited to a myopic view considering only past or present information. To account for future impacts, we propose a stochastic sequential decision-making framework for filtering networked data with a policy that adapts filtering to expanding graphs. By representing filter shifts as agents, we model the filter as a multi-agent system and train the policy following multi-agent reinforcement learning. This accounts for long-term rewards and captures expansion dynamics through sequential decision-making. Moreover, we develop a context-aware graph neural network to parameterize the policy, which tunes filter parameters based on information of both the graph and agents. Experiments on synthetic and real datasets from cold-start recommendation to COVID prediction highlight the benefits of using a sequential decision-making perspective over batch and online filtering alternatives.
【5】Exploring Subnetwork Interactions in Heterogeneous Brain Network via Prior-Informed Graph Learning
标题：通过先验图学习探索异类脑网络中的子网络相互作用
链接：https://arxiv.org/abs/2603.19307
摘要：功能子网络之间复杂的相互作用建模对于精神障碍的诊断和功能通路的识别至关重要。然而，由于训练样本的数量有限，学习底层子网络的相互作用对于现有的基于transformer的方法仍然是一个重大挑战。为了解决这些挑战，我们提出了KD-Brain，一个先验知识图学习框架，用于明确编码先验知识来指导学习过程。具体来说，我们设计了一个语义条件交互机制，将语义先验注入到注意查询，明确导航子网的交互功能的基础上的身份。此外，我们引入了病理一致性约束，通过将学习到的相互作用分布与临床先验进行对齐来规范模型优化。此外，KD-Brain在广泛的疾病诊断任务上具有最先进的性能，并识别出与精神病病理生理学一致的可解释的生物标志物。我们的代码可在https://anonymous.4open.science/r/KDBrain上获得。
摘要：Modeling the complex interactions among functional subnetworks is crucial for the diagnosis of mental disorders and the identification of functional pathways. However, learning the interactions of the underlying subnetworks remains a significant challenge for existing Transformer-based methods due to the limited number of training samples. To address these challenges, we propose KD-Brain, a Prior-Informed Graph Learning framework for explicitly encoding prior knowledge to guide the learning process. Specifically, we design a Semantic-Conditioned Interaction mechanism that injects semantic priors into the attention query, explicitly navigating the subnetwork interactions based on their functional identities. Furthermore, we introduce a Pathology-Consistent Constraint, which regularizes the model optimization by aligning the learned interaction distributions with clinical priors. Additionally, KD-Brain leads to state-of-the-art performance on a wide range of disorder diagnosis tasks and identifies interpretable biomarkers consistent with psychiatric pathophysiology. Our code is available at https://anonymous.4open.science/r/KDBrain.
【6】Graph-Informed Adversarial Modeling: Infimal Subadditivity of Interpolative Divergences
标题：图知情对抗建模：插值分歧的下半可加性
链接：https://arxiv.org/abs/2603.20025
备注：32 pages, 9 figures
摘要：我们研究了当目标分布根据已知贝叶斯网络分解时的对抗学习。对于插值发散，包括$（f，Γ）$-发散，我们证明了一个新的无穷次可加性原理，证明了在适当的条件下，全局变分偏差由与图对齐的族级偏差的平均值控制.在加法体系中，这个替代项是精确的。这为用具有局部族级鉴别器的图形通知GAN替换具有单片鉴别器的图形不可知GAN提供了变化的理由。结果不需要优化器本身根据图进行因式分解。我们还获得了并行的结果，积分概率度量和近端最佳运输分歧，确定自然类的理论适用，目前的实验表明，改善稳定性和结构恢复相对于图形不可知的基线。
摘要：We study adversarial learning when the target distribution factorizes according to a known Bayesian network. For interpolative divergences, including $(f,Γ)$-divergences, we prove a new infimal subadditivity principle showing that, under suitable conditions, a global variational discrepancy is controlled by an average of family-level discrepancies aligned with the graph. In an additive regime, this surrogate is exact. This provides a variational justification for replacing a graph-agnostic GAN with a monolithic discriminator by a graph-informed GAN with localized family-level discriminators. The result does not require the optimizer itself to factorize according to the graph. We also obtain parallel results for integral probability metrics and proximal optimal transport divergences, identify natural discriminator classes for which the theory applies, and present experiments showing improved stability and structural recovery relative to graph-agnostic baselines.
【7】Modeling subgrid scale production rates on complex meshes using graph neural networks
标题：基于图神经网络的复杂网格下亚网格规模生产率建模
链接：https://arxiv.org/abs/2603.19841
摘要：大涡模拟（LES）需要关闭过滤生产率，因为解决领域不包含所有的相关性，管理化学源项。我们开发了一个图形神经网络（GNN），预测过滤后的物质生产率的非均匀网格过滤后的质量分数和温度的输入。直接数值模拟的湍流预混氢-甲烷射流火焰与氢的分数为10%，50%和80%提供的数据集。所有的字段Favre过滤与过滤器宽度匹配的操作网格，并学习从网格点连接构造的子域图上进行。使用一组紧凑的反应物、中间体和产物，并且它们的过滤的生产率形成目标。该模型在10%和80%的混合上进行训练，并在看不见的50%混合上进行评估，以测试交叉合成泛化。GNN与评估滤波状态下速率的未闭合参考以及需要重新划分网格的卷积神经网络基线进行比较。在分布内和分布外的情况下，GNN产生较低的误差和更接近的统计协议与参考数据。此外，该模型表现出强大的泛化在不同的滤波器宽度，而无需重新训练，保持有界误差在较粗的空间分辨率。一个面向后的步骤配置进一步证实了预测效果的实际相关的几何形状。这些结果突出了GNNs作为复杂网格上LES的强大数据驱动闭合模型的能力。
摘要 ：Large-eddy simulations (LES) require closures for filtered production rates because the resolved fields do not contain all correlations that govern chemical source terms. We develop a graph neural network (GNN) that predicts filtered species production rates on non-uniform meshes from inputs of filtered mass fractions and temperature. Direct numerical simulations of turbulent premixed hydrogen-methane jet flames with hydrogen fractions of 10%, 50%, and 80% provide the dataset. All fields are Favre filtered with the filter width matched to the operating mesh, and learning is performed on subdomain graphs constructed from mesh-point connectivity. A compact set of reactants, intermediates, and products is used, and their filtered production rates form the targets. The model is trained on 10% and 80% blends and evaluated on the unseen 50% blend to test cross-composition generalization. The GNN is compared against an unclosed reference that evaluates rates at the filtered state, and a convolutional neural network baseline that requires remeshing. Across in-distribution and out-of-distribution cases, the GNN yields lower errors and closer statistical agreement with the reference data. Furthermore, the model demonstrates robust generalization across varying filter widths without retraining, maintaining bounded errors at coarser spatial resolutions. A backward facing step configuration further confirms prediction efficacy on a practically relevant geometry. These results highlight the capability of GNNs as robust data-driven closure models for LES on complex meshes.
【8】Subspace Projection Methods for Fast Spectral Embeddings of Evolving Graphs
标题：演化图快速谱嵌入的子空间投影方法
链接：https://arxiv.org/abs/2603.19439
摘要：一些图数据挖掘、信号处理和机器学习下游任务依赖于与关联邻接或拉普拉斯矩阵的特征向量相关的信息。经典的特征分解方法是强大的，当矩阵保持静态，但不能适用于矩阵的条目更新或行和列的数量频繁增加的问题。当图动态变化并且边和/或节点被添加和移除时，这样的场景经常发生在图分析中。本文提出了一种新的算法框架来更新与初始邻接矩阵或拉普拉斯矩阵的前导特征值相关联的特征向量，随着图的动态演化。该算法基于Rayleigh-Ritz投影，将原特征值问题投影到一个约束子空间上，该子空间理想地封装了与所求特征向量相关的不变子空间。根据特征向量扰动分析的思想，我们提出了一种新的方法来建立投影子空间。所提出的框架具有较低的计算和内存的复杂性相对于竞争的替代品，而实证结果显示出强大的定性性能，无论是在特征向量近似和准确性的下游学习任务的中心节点识别和节点聚类。
摘要：Several graph data mining, signal processing, and machine learning downstream tasks rely on information related to the eigenvectors of the associated adjacency or Laplacian matrix. Classical eigendecomposition methods are powerful when the matrix remains static but cannot be applied to problems where the matrix entries are updated or the number of rows and columns increases frequently. Such scenarios occur routinely in graph analytics when the graph is changing dynamically and either edges and/or nodes are being added and removed. This paper puts forth a new algorithmic framework to update the eigenvectors associated with the leading eigenvalues of an initial adjacency or Laplacian matrix as the graph evolves dynamically. The proposed algorithm is based on Rayleigh-Ritz projections, in which the original eigenvalue problem is projected onto a restricted subspace which ideally encapsulates the invariant subspace associated with the sought eigenvectors. Following ideas from eigenvector perturbation analysis, we present a new methodology to build the projection subspace. The proposed framework features lower computational and memory complexity with respect to competitive alternatives while empirical results show strong qualitative performance, both in terms of eigenvector approximation and accuracy of downstream learning tasks of central node identification and node clustering.
Transformer(5篇)
【1】On the Ability of Transformers to Verify Plans
标题：论Transformer验证计划的能力
链接：https://arxiv.org/abs/2603.19954
摘要：Transformers在人工智能规划任务中表现出不一致的成功，并且对何时应该预期泛化的理论理解有限。我们通过分析仅解码器模型的能力来验证给定计划是否正确解决给定规划实例，从而采取重要步骤来解决这一差距。要分析的一般设置中的对象的数量-从而有效的输入字母-在测试时增长，我们引入C*-RASP，C-RASP的扩展，旨在建立长度泛化保证下的同时增长的序列长度和词汇量的Transformers器。我们的研究结果确定了一个大类的经典规划域，Transformers可以证明学习验证长计划，和结构特性，显着影响学习的长度概括的解决方案。实验证实了我们的理论。
摘要：Transformers have shown inconsistent success in AI planning tasks, and theoretical understanding of when generalization should be expected has been limited. We take important steps towards addressing this gap by analyzing the ability of decoder-only models to verify whether a given plan correctly solves a given planning instance. To analyse the general setting where the number of objects -- and thus the effective input alphabet -- grows at test time, we introduce C*-RASP, an extension of C-RASP designed to establish length generalization guarantees for transformers under the simultaneous growth in sequence length and vocabulary size. Our results identify a large class of classical planning domains for which transformers can provably learn to verify long plans, and structural properties that significantly affects the learnability of length generalizable solutions. Empirical experiments corroborate our theory.
【2】Dual Path Attribution: Efficient Attribution for SwiGLU-Transformers through Layer-Wise Target Propagation
标题：双路径属性：通过分层目标传播实现SwiGLU变形器的有效属性
链接：https://arxiv.org/abs/2603.19742
摘要：理解基于transformer的大型语言模型（LLM）的内部机制对于其可靠部署和有效运行至关重要。虽然最近的努力已经产生了过多的归因方法，试图平衡的忠实性和计算效率，密集的组件归因仍然是昂贵的。在这项工作中，我们介绍了双路径属性（DPA），一种新的框架，忠实地跟踪信息流冻结Transformer在一个向前和一个向后的通行证，而不需要反事实的例子。DPA将SwiGLU Transformers的计算结构解析分解并线性化为不同的路径，沿着这些路径，DPA传播目标解嵌入向量以在每个残差位置处接收有效表示。这种以目标为中心的传播在模型组件数量、扩展到长输入序列和密集组件属性方面实现了O（1）的时间复杂度。对标准可解释性基准的大量实验表明，与现有基准相比，DPA实现了最先进的忠实性和前所未有的效率。
摘要：Understanding the internal mechanisms of transformer-based large language models (LLMs) is crucial for their reliable deployment and effective operation. While recent efforts have yielded a plethora of attribution methods attempting to balance faithfulness and computational efficiency, dense component attribution remains prohibitively expensive. In this work, we introduce Dual Path Attribution (DPA), a novel framework that faithfully traces information flow on the frozen transformer in one forward and one backward pass without requiring counterfactual examples. DPA analytically decomposes and linearizes the computational structure of the SwiGLU Transformers into distinct pathways along which it propagates a targeted unembedding vector to receive the effective representation at each residual position. This target-centric propagation achieves O(1) time complexity with respect to the number of model components, scaling to long input sequences and dense component attribution. Extensive experiments on standard interpretability benchmarks demonstrate that DPA achieves state-of-the-art faithfulness and unprecedented efficiency compared to existing baselines.
【3】The Residual Stream Is All You Need: On the Redundancy of the KV Cache in Transformer Inference
标题：剩余流即可满足您的需求：关于Transformer推理中KV缓存的冗余性
链接：https://arxiv.org/abs/2603.19664
备注：14
摘要 ：键值（KV）缓存被广泛视为Transformer推理中的基本状态，并且大量的工作工程师策略来压缩、驱逐或近似其条目。我们证明了这种状态是完全冗余的：每一层的键和值都是残差流的确定性投影，并且从每个令牌的单个残差向量重新计算它们会导致零重构误差，不是近似的，而是位相同的。我们验证了这六个模型从四个架构系列（135 M到4 B参数）。在每一层的跨任务剩余修补产生修补和原始输出分布之间的D_KL = 0，确认剩余流满足马尔可夫性质并且是唯一的信息携带状态。在所有测试模型上，完全删除缓存并从头开始重新计算会在贪婪解码下产生令牌相同的输出。我们使用KV-Direct建立了这个结果，这是一种有界内存推理方案，它检查剩余向量（Gemma 3- 4 B上每个令牌5 KB）而不是完整的KV对（136 KB），按需重新计算键和值。超过20个会话回合，KV直接保持峰值内存在42 MB，而标准缓存增长超过103 MB。针对五个回收基线（H2O，StreamingLLM，SnapKV，TOVA，仅窗口），KV-Direct在每个缓存预算中都保持100%的令牌匹配;所有基线都降低到5- 28%。每个操作的延迟分析显示，在中等批量大小下，重新计算的速度比读取缓存张量快5倍。代码可在https://github.com/Kaleemullahqasim/KV-Direct上获得。
摘要：The key-value (KV) cache is widely treated as essential state in transformer inference, and a large body of work engineers policies to compress, evict, or approximate its entries. We prove that this state is entirely redundant: keys and values at every layer are deterministic projections of the residual stream, and recomputing them from a single residual vector per token incurs exactly zero reconstruction error, not approximately, but bit-identically. We verify this across six models from four architecture families (135M to 4B parameters). Cross-task residual patching at every layer produces D_KL = 0 between patched and original output distributions, confirming that the residual stream satisfies a Markov property and is the sole information-carrying state. Removing the cache entirely and recomputing from scratch yields token-identical output under greedy decoding on all models tested. We build on this result with KV-Direct, a bounded-memory inference scheme that checkpoints residual vectors (5 KB per token on Gemma 3-4B) instead of full KV pairs (136 KB), recomputing keys and values on demand. Over 20 conversation turns, KV-Direct holds peak memory at 42 MB while the standard cache grows past 103 MB. Against five eviction baselines (H2O, StreamingLLM, SnapKV, TOVA, window-only), KV-Direct maintains 100% token match at every cache budget; all baselines degrade to 5-28%. A per-operation latency analysis shows recomputation runs up to 5x faster than reading cached tensors at moderate batch sizes. Code is available at https://github.com/Kaleemullahqasim/KV-Direct.
【4】DAPA: Distribution Aware Piecewise Activation Functions for On-Device Transformer Inference and Training
标题：DAPA：用于设备上Transformer推理和训练的分布感知分段激活功能
链接：https://arxiv.org/abs/2603.19338
摘要：非线性激活函数在设备上的推理和训练中起着关键作用，因为它们不仅消耗大量的硬件资源，而且对系统性能和能源效率产生重大影响。在这项工作中，我们提出了分布感知分段激活（DAPA），一个可微的和硬件友好的激活功能的Transformer架构，通过利用预激活数据的分布。DAPA采用非均匀分段近似，将更精细的片段分配给分布的高概率区域，从而提高了先前分段线性方法的可推广性。使用分布加权均方误差进一步量化得到的近似值，以减少硬件部署的延迟和资源利用率。我们的HLS实现表明，DAPA将GELU计算速度提高了16倍，将DSP利用率降低了16倍，同时在Vision Transformers和GPT-2模型之间保持相当或更好的性能。
摘要：Non-linear activation functions play a pivotal role in on-device inference and training, as they not only consume substantial hardware resources but also impose a significant impact on system performance and energy efficiency. In this work, we propose Distribution-Aware Piecewise Activation (DAPA), a differentiable and hardware-friendly activation function for Transformer architectures by exploiting the distribution of pre-activation data. DAPA employs a non-uniform piecewise approximation that allocates finer segments to high-probability regions of the distribution, improving generalizability over prior piecewise linear methods. The resulting approximation is further quantized using Distribution-Weighted Mean Square Error to reduce latency and resource utilization for hardware deployment. Our HLS implementation demonstrates that DAPA speeds up GELU computation by 16$\times$ and decreases DSP utilization by 16$\times$ while maintaining comparable or better performance across vision Transformers and GPT-2 models.
【5】Transformers are Stateless Differentiable Neural Computers
标题：Transformer是无状态可区分的神经计算机
链接：https://arxiv.org/abs/2603.19272
备注：7 pages
摘要：微分神经计算机（DNC）是作为循环架构引入的，它配备了支持微分读写操作的可寻址外部存储器。相比之下，Transformers名义上是基于多头自关注的前馈架构。在这项工作中，我们给出了一个正式的推导表明，因果Transformer层正是一个无状态的可微分神经计算机（sDNC），其中（1）控制器没有经常性的内部状态，（2）外部存储器是一个写一次矩阵的值向量，（3）基于内容的寻址通过键实现注意力，（4）多头注意力对应于多个并行读取头。我们进一步将这种等价性扩展到交叉注意，表明编码器-解码器Transformers正是具有不同读取和写入存储器的SDNC。我们的研究结果提供了一个统一的内存为中心的解释Transformers，并有助于正在进行的努力，把现代大型语言模型的原则计算框架。
摘要：Differentiable Neural Computers (DNCs) were introduced as recurrent architectures equipped with an addressable external memory supporting differentiable read and write operations. Transformers, in contrast, are nominally feedforward architectures based on multi-head self-attention. In this work we give a formal derivation showing that a causal Transformer layer is exactly a stateless Differentiable Neural Computer (sDNC) where (1) the controller has no recurrent internal state, (2) the external memory is a write-once matrix of value vectors, (3) content-based addressing via keys implements attention, and (4) multi-head attention corresponds to multiple parallel read heads. We further extend this equivalence to cross-attention, showing that encoder-decoder Transformers are precisely sDNCs with distinct read-from and write-to memories. Our results provide a unified memory-centric interpretation of Transformers and contribute to the ongoing effort to place modern large language models in a principled computational framework.
GAN|对抗|攻击|生成相关(6篇)
【1】Conditioning Protein Generation via Hopfield Pattern Multiplicity
标题：通过Hopfield模式多重性产生条件蛋白质
链接：https://arxiv.org/abs/2603.20115
摘要：通过随机注意生成蛋白质序列可以在没有训练的情况下从小的比对中产生看似合理的家族成员，但是平等地对待所有存储的序列，并且不能直接生成感兴趣的功能子集。我们发现，一个单一的标量参数，作为一个偏差添加到采样器的注意力的logits，不断转移一代从完整的家庭向用户指定的子集，没有再培训和模型架构没有变化。从业者提供一小部分序列（例如，来自结合筛选的命中）和多重性比率，该比率控制生成对它们有利的程度。该方法不知道子集代表什么：结合、稳定性、特异性或任何其他属性。我们发现，条件是准确的采样器的内部表示的水平，但解码的序列表型可以短，因为用于编码序列的降维并不总是保持残基水平的变化，定义的功能分裂。我们把这种差异的校准差距，并表明，它是预测一个简单的几何测量以及如何编码分离的功能子集从其余的家庭。五个Pfam家族（Kunitz，SH 3，WW，Homeobox和Forkhead域）的实验证实了在四倍范围的几何形状之间的单调关系的分离和间隙。应用于靶向参与疼痛信号传导的钙通道的ω-芋螺毒素肽，来自23种特征结合剂的策展播种产生超过一千种候选物，其保留主要药效团和所有实验鉴定的结合决定簇。这些结果表明，随机注意使从业者能够将少数实验表征的序列扩展到不同的候选库中，而无需重新训练生成模型。
摘要 ：Protein sequence generation via stochastic attention produces plausible family members from small alignments without training, but treats all stored sequences equally and cannot direct generation toward a functional subset of interest. We show that a single scalar parameter, added as a bias to the sampler's attention logits, continuously shifts generation from the full family toward a user-specified subset, with no retraining and no change to the model architecture. A practitioner supplies a small set of sequences (for example, hits from a binding screen) and a multiplicity ratio that controls how strongly generation favors them. The method is agnostic to what the subset represents: binding, stability, specificity, or any other property. We find that the conditioning is exact at the level of the sampler's internal representation, but that the decoded sequence phenotype can fall short because the dimensionality reduction used to encode sequences does not always preserve the residue-level variation that defines the functional split. We term this discrepancy the calibration gap and show that it is predicted by a simple geometric measure of how well the encoding separates the functional subset from the rest of the family. Experiments on five Pfam families (Kunitz, SH3, WW, Homeobox, and Forkhead domains) confirm the monotonic relationship between separation and gap across a fourfold range of geometries. Applied to omega-conotoxin peptides targeting a calcium channel involved in pain signaling, curated seeding from 23 characterized binders produces over a thousand candidates that preserve the primary pharmacophore and all experimentally identified binding determinants. These results show that stochastic attention enables practitioners to expand a handful of experimentally characterized sequences into diverse candidate libraries without retraining a generative model.
【2】Channel Prediction-Based Physical Layer Authentication under Consecutive Spoofing Attacks
标题：连续欺骗攻击下基于通道预测的物理层认证
链接：https://arxiv.org/abs/2603.19962
摘要：无线网络极易受到欺骗攻击，特别是当攻击者发送连续的欺骗数据包时。传统的物理层认证（PLA）方法大多集中在单包欺骗攻击。然而，在连续的欺骗攻击下，由于设备移动性和信道衰落引起的信道演进，它们变得无效。为了应对这一挑战，我们提出了一个基于信道预测的PLA框架。具体地，采用基于变换器的信道预测模块来预测欺骗间隔期间的合法CSI测量，并且基于认证决定用预测或观察到的CSI测量自适应地更新信道预测模块的输入，以确保针对持续欺骗的鲁棒性。在瑞利衰落信道下的仿真结果表明，该方法具有较低的预测误差和较高的认证精度，即使在扩展欺骗攻击下也具有较好的鲁棒性。
摘要：Wireless networks are highly vulnerable to spoofing attacks, especially when attackers transmit consecutive spoofing packets. Conventional physical layer authentication (PLA) methods have mostly focused on single-packet spoofing attack. However, under consecutive spoofing attacks, they become ineffective due to channel evolution caused by device mobility and channel fading. To address this challenge, we propose a channel prediction-based PLA framework. Specifically, a Transformer-based channel prediction module is employed to predict legitimate CSI measurements during spoofing interval, and the input of channel prediction module is adaptively updated with predicted or observed CSI measurements based on the authentication decision to ensure robustness against sustained spoofing. Simulation results under Rayleigh fading channels demonstrate that the proposed approach achieves low prediction error and significantly higher authentication accuracy than conventional benchmark, maintaining robustness even under extended spoofing attacks.
【3】ARMOR: Adaptive Resilience Against Model Poisoning Attacks in Continual Federated Learning for Mobile Indoor Localization
标题：ARMOR：针对移动室内定位的连续联邦学习中模型中毒攻击的自适应弹性
链接：https://arxiv.org/abs/2603.19594
摘要：室内定位对于从资产跟踪到提供个性化服务的应用越来越重要。联邦学习（FL）通过使用来自移动设备的分布式数据训练集中式全局模型（GM），而无需共享原始数据，从而提供了一种隐私保护方法。然而，现实世界的部署需要持续的联邦学习（CFL）设置，其中GM在设备异构性和不断变化的室内环境下接收持续更新。在这种动态条件下，错误或有偏见的更新可能会导致GM偏离其预期的学习轨迹，逐渐降低内部GM表示和GM本地化性能。对抗性模型中毒攻击进一步加剧了这种脆弱性。为了应对这一挑战，我们提出了ARMOR，这是一个基于CFL的新框架，可以在不断更新的过程中监控和保护GM。ARMOR引入了一种新的状态空间模型（SSM），该模型学习GM权重张量的历史演变，并预测GM权重张量的预期下一个状态。通过将传入的本地更新与此SSM投影进行比较，ARMOR检测偏差并在本地更新与GM聚合之前选择性地减轻损坏的更新。该机制能够鲁棒地适应时间环境动态，并在防止GM损坏的同时减轻模型中毒攻击的影响。在真实世界条件下的实验评估表明，ARMOR实现了显着的改进，与最先进的室内定位框架相比，平均误差减少了8.0倍，最坏情况下的误差减少了4.97倍，证明了对使用真实世界数据和移动设备测试的模型损坏的强大弹性。
摘要：Indoor localization has become increasingly essential for applications ranging from asset tracking to delivering personalized services. Federated learning (FL) offers a privacy-preserving approach by training a centralized global model (GM) using distributed data from mobile devices without sharing raw data. However, real-world deployments require a continual federated learning (CFL) setting, where the GM receives continual updates under device heterogeneity and evolving indoor environments. In such dynamic conditions, erroneous or biased updates can cause the GM to deviate from its expected learning trajectory, gradually degrading internal GM representations and GM localization performance. This vulnerability is further exacerbated by adversarial model poisoning attacks. To address this challenge, we propose ARMOR, a novel CFL-based framework that monitors and safeguards the GM during continual updates. ARMOR introduces a novel state-space model (SSM) that learns the historical evolution of GM weight tensors and predicts the expected next state of weight tensors of the GM. By comparing incoming local updates with this SSM projection, ARMOR detects deviations and selectively mitigates corrupted updates before local updates are aggregated with the GM. This mechanism enables robust adaptation to temporal environmental dynamics and mitigate the effects of model poisoning attacks while preventing GM corruption. Experimental evaluations in real-world conditions indicate that ARMOR achieves notable improvements, with up to 8.0x reduction in mean error and 4.97x reduction in worst-case error compared to state-of-the-art indoor localization frameworks, demonstrating strong resilience against model corruption tested using real-world data and mobile devices.
【4】Warm-Start Flow Matching for Guaranteed Fast Text/Image Generation
标题：热启动流匹配，保证快速文本/图像生成
链接：https://arxiv.org/abs/2603.19360
摘要：当前的自回归（AR）LLM、基于扩散的文本/图像生成模型和最近的流匹配（FM）算法能够生成优质的文本/图像样本。然而，这些模型中的推理或样本生成通常非常耗时且计算要求很高，这主要是由于与令牌长度或扩散步骤数量相对应的大量函数评估。这也需要大量的GPU资源、时间和电力。在这项工作中，我们提出了一种新的解决方案，以减少流匹配算法的样本生成时间的保证加速因子，而不牺牲所生成的样本的质量。我们的主要思想是利用计算轻量级生成模型，其生成时间与目标AR/FM模型相比可以忽略不计。该算法将来自轻量级模型的草稿样本作为FM算法的初始分布，该样本质量不理想，但生成速度快。与采用纯噪声的FM的常规使用不同（例如，高斯或均匀）初始分布时，草稿样本已经具有良好的质量，因此我们可以将开始时间设置为更接近结束时间，而不是在纯噪声FM情况下设置为0。这将显著减少达到目标数据分布所需的时间步数，并保证了加速因子。我们的想法被称为Warm-Start FM或WS-FM，本质上可以被视为从低质量草稿样本到高质量样本的生成模型。作为概念证明，我们在一些合成玩具数据以及现实世界的文本和图像生成任务上展示了这个想法，说明我们的想法在不牺牲生成样本质量的情况下保证了样本生成的速度。
摘要 ：Current auto-regressive (AR) LLMs, diffusion-based text/image generative models, and recent flow matching (FM) algorithms are capable of generating premium quality text/image samples. However, the inference or sample generation in these models is often very time-consuming and computationally demanding, mainly due to large numbers of function evaluations corresponding to the lengths of tokens or the numbers of diffusion steps. This also necessitates heavy GPU resources, time, and electricity. In this work we propose a novel solution to reduce the sample generation time of flow matching algorithms by a guaranteed speed-up factor, without sacrificing the quality of the generated samples. Our key idea is to utilize computationally lightweight generative models whose generation time is negligible compared to that of the target AR/FM models. The draft samples from a lightweight model, whose quality is not satisfactory but fast to generate, are regarded as an initial distribution for a FM algorithm. Unlike conventional usage of FM that takes a pure noise (e.g., Gaussian or uniform) initial distribution, the draft samples are already of decent quality, so we can set the starting time to be closer to the end time rather than 0 in the pure noise FM case. This will significantly reduce the number of time steps to reach the target data distribution, and the speed-up factor is guaranteed. Our idea, dubbed {\em Warm-Start FM} or WS-FM, can essentially be seen as a {\em learning-to-refine} generative model from low-quality draft samples to high-quality samples. As a proof of concept, we demonstrate the idea on some synthetic toy data as well as real-world text and image generation tasks, illustrating that our idea offers guaranteed speed-up in sample generation without sacrificing the quality of the generated samples.
【5】A Novel Solution for Zero-Day Attack Detection in IDS using Self-Attention and Jensen-Shannon Divergence in WGAN-GP
标题：WGAN-GP中利用自我注意力和Jensen-Shannon分歧进行IDS零日攻击检测的新解决方案
链接：https://arxiv.org/abs/2603.19350
备注：40 pages, 5 figures, including references
摘要：日益复杂的网络威胁，特别是零日攻击，对网络安全构成了重大挑战。零日攻击利用未知的漏洞，使其难以检测和防御。现有的方法修补漏洞并部署入侵检测系统（IDS）。本文提出了一种新的基于梯度惩罚的Wasserstein GANs（WGAN-GP）网络流量合成方法，该方法能够模拟零日模式，丰富数据多样性，提高入侵检测系统的泛化能力。首先介绍了SA-WGAN-GP，它增加了一个自注意（SA）机制，通过密集投影后将特征向量重塑为令牌来捕获长范围的跨特征依赖关系。然后提出了一个JS-WGAN-GP，它添加了一个基于Jensen-Shannon（JS）发散的辅助训练器，该辅助训练器使用二进制交叉熵（BCE）进行训练，在更新期间冻结，并用于正则化生成器以获得更平滑的梯度和更高的样本质量。第三，通过将SA机制与JS发散相结合，创建了SA-JS-WGAN-GP，从而增强了WGAN-GP的数据生成能力。由于数据增强并不等同于真正的零日攻击发现，我们通过NSL-KDD数据集上的leave-one-attack-type-out方法来模拟零日攻击，以训练所有GANs和IDS模型，评估所提出的解决方案的有效性。评估结果表明，将SA和JS分歧集成到WGAN-GP中可以获得更好的IDS性能和更有效的零日风险检测。
摘要：The increasing sophistication of cyber threats, especially zero-day attacks, poses a significant challenge to cybersecurity. Zero-day attacks exploit unknown vulnerabilities, making them difficult to detect and defend against. Existing approaches patch flaws and deploy an Intrusion Detection System (IDS). Using advanced Wasserstein GANs with Gradient Penalty (WGAN-GP), this paper makes a novel proposition to synthesize network traffic that mimics zero-day patterns, enriching data diversity and improving IDS generalization. SA-WGAN-GP is first introduced, which adds a Self-Attention (SA) mechanism to capture long-range cross-feature dependencies by reshaping the feature vector into tokens after dense projections. A JS-WGAN-GP is then proposed, which adds a Jensen-Shannon (JS) divergence-based auxiliary discriminator that is trained with Binary Cross-Entropy (BCE), frozen during updates, and used to regularize the generator for smoother gradients and higher sample quality. Third, SA-JS-WGAN-GP is created by combining the SA mechanism with JS divergence, thereby enhancing the data generation ability of WGAN-GP. As data augmentation does not equate with true zero-day attack discovery, we emulate zero-day attacks via the leave-one-attack-type-out method on the NSL-KDD dataset for training all GANs and IDS models in the assessment of the effectiveness of the proposed solution. The evaluation results show that integrating SA and JS divergence into WGAN-GP yields superior IDS performance and more effective zero-day risk detection.
【6】Exploring the Agentic Frontier of Verilog Code Generation
标题：探索Verilog代码生成的超级前沿
链接：https://arxiv.org/abs/2603.19347
摘要：大型语言模型（LLM）在Python和C++等流行语言的代码生成方面取得了快速进展。这些最近的收益中有许多可以归因于使用“代理”，这些代理将领域相关工具与LLM一起包装。近年来，Verilog等硬件设计语言的代码生成也得到了改进，但代理框架对Verilog代码生成任务的影响仍不清楚。在这项工作中，我们提出了第一个系统的评估代理LLM的Verilog生成，使用最近推出的CVDP基准。我们还介绍了几个开源的硬件设计代理线束，为未来的工作提供了一个模型无关的基线。通过跨前沿模型的对照实验，我们研究了结构化提示和工具设计如何影响性能，分析代理失败模式和工具使用模式，比较开源和闭源模型，并提供成功和失败的代理运行的定性例子。我们的研究结果表明，围绕前沿模型的天真代理会降低性能（相对于具有优化提示的标准前向传递），但结构化线束有意义地匹配，并在某些情况下超过非代理基线。我们发现开源和闭源模型之间的性能差距是由更高的崩溃率和更弱的工具输出解释驱动的。我们的探索照亮了未来设计专用代理的verilog生成的路径。
摘要：Large language models (LLMs) have made rapid advancements in code generation for popular languages such as Python and C++. Many of these recent gains can be attributed to the use of ``agents'' that wrap domain-relevant tools alongside LLMs. Hardware design languages such as Verilog have also seen improved code generation in recent years, but the impact of agentic frameworks on Verilog code generation tasks remains unclear. In this work, we present the first systematic evaluation of agentic LLMs for Verilog generation, using the recently introduced CVDP benchmark. We also introduce several open-source hardware design agent harnesses, providing a model-agnostic baseline for future work. Through controlled experiments across frontier models, we study how structured prompting and tool design affect performance, analyze agent failure modes and tool usage patterns, compare open-source and closed-source models, and provide qualitative examples of successful and failed agent runs. Our results show that naive agentic wrapping around frontier models can degrade performance (relative to standard forward passes with optimized prompts), but that structured harnesses meaningfully match and in some cases exceed non-agentic baselines. We find that the performance gap between open and closed source models is driven by both higher crash rates and weaker tool output interpretation. Our exploration illuminates the path towards designing special-purpose agents for verilog generation in the future.
半/弱/无/有监督|不确定性|主动学习(4篇)
【1】Var-JEPA: A Variational Formulation of the Joint-Embedding Predictive Architecture -- Bridging Predictive and Generative Self-Supervised Learning
标题：Var-JEPA：联合嵌入预测架构的变分公式--连接预测和生成性自我监督学习
链接：https://arxiv.org/abs/2603.20111
摘要：联合嵌入预测架构（Joint-Embedding Predictive Architecture，JEPA）通常被视为基于可能性的自监督学习的非生成替代方案，强调表示空间中的预测，而不是观察空间中的重建。我们认为，由此产生的分离概率生成模型主要是修辞而不是结构：典型的JEPA设计，将编码器与上下文到目标预测器耦合，反映了当变分推理应用于特定类别的耦合潜变量模型时获得的变分后验和学习的条件先验，标准JEPA可以被看作是一种确定性的专门化，其中正则化是通过架构和训练算法而不是显式的可能性来施加的。在此基础上，我们推导出变分JEPA（Var-JEPA），它通过优化单个证据下限（ELBO）来明确潜在的生成结构。这产生了有意义的表示，而无需专门的抗崩溃正则化器，并允许在潜在空间中进行原则性的不确定性量化。我们实例化了表格数据框架（Var-T-JEPA），并实现了强大的表示学习和下游性能，在T-JEPA上不断改进，同时与强大的原始特征基线保持竞争力。
摘要：The Joint-Embedding Predictive Architecture (JEPA) is often seen as a non-generative alternative to likelihood-based self-supervised learning, emphasizing prediction in representation space rather than reconstruction in observation space. We argue that the resulting separation from probabilistic generative modeling is largely rhetorical rather than structural: the canonical JEPA design, coupled encoders with a context-to-target predictor, mirrors the variational posteriors and learned conditional priors obtained when variational inference is applied to a particular class of coupled latent-variable models, and standard JEPA can be viewed as a deterministic specialization in which regularization is imposed via architectural and training heuristics rather than an explicit likelihood. Building on this view, we derive the Variational JEPA (Var-JEPA), which makes the latent generative structure explicit by optimizing a single Evidence Lower Bound (ELBO). This yields meaningful representations without ad-hoc anti-collapse regularizers and allows principled uncertainty quantification in the latent space. We instantiate the framework for tabular data (Var-T-JEPA) and achieve strong representation learning and downstream performance, consistently improving over T-JEPA while remaining competitive with strong raw-feature baselines.
【2】Ontology-Based Knowledge Modeling and Uncertainty-Aware Outdoor Air Quality Assessment Using Weighted Interval Type-2 Fuzzy Logic
标题：基于本体的知识建模及基于加权区间2型模糊逻辑的不确定性室外空气质量评价
链接：https://arxiv.org/abs/2603.19683
摘要：室外空气污染是环境和公众健康的一个主要问题，特别是在城市化迅速发展的地区。印度空气质量指数（IND-AQI）由中央污染控制委员会（CPCB）开发，是一个标准化的空气质量报告系统，基于PM2.5，PM10），二氧化氮（NO2），二氧化硫（SO2），臭氧（O3），一氧化碳（CO）和氨（NH3）等污染物。然而，AQI的传统计算使用明确的阈值和确定性聚合规则，这不适合处理不确定性和类之间的转换。为了解决这些局限性，本研究提出了一个混合的基于本体的不确定性感知框架集成加权区间2型模糊逻辑与语义知识建模。区间型2模糊集用于模拟AQI类边界附近的不确定性，而污染物的重要性权重确定使用区间型2模糊层次分析法（IT 2-FAHP），以反映其相对健康影响。此外，基于OWL的空气质量本体扩展语义传感器网络（SSN）本体的开发，以表示污染物，监测站，AQI类别，监管标准和环境治理行动。语义推理使用SWRL规则实现，并通过SPARQL查询进行验证，以推断AQI类别，健康风险和建议的缓解措施。使用CPCB空气质量数据集的实验评估表明，与传统的清晰和1型模糊方法相比，所提出的框架提高了AQI分类的可靠性和不确定性处理，同时为空气质量监测系统提供了可解释的语义推理和智能决策支持
摘要：Outdoor air pollution is a major concern for the environment and public health, especially in areas where urbanization is taking place rapidly. The Indian Air Quality Index (IND-AQI), developed by the Central Pollution Control Board (CPCB), is a standardized reporting system for air quality based on pollutants such as PM2.5, PM10), nitrogen dioxide (NO2), sulfur dioxide (SO2), ozone (O3), carbon monoxide (CO), and ammonia (NH3). However, the traditional calculation of the AQI uses crisp thresholds and deterministic aggregation rules, which are not suitable for handling uncertainty and transitions between classes. To address these limitations, this study proposes a hybrid ontology-based uncertainty-aware framework integrating Weighted Interval Type-2 Fuzzy Logic with semantic knowledge modeling. Interval Type-2 fuzzy sets are used to model uncertainty near AQI class boundaries, while pollutant importance weights are determined using Interval Type-2 Fuzzy Analytic Hierarchy Process (IT2-FAHP) to reflect their relative health impacts. In addition, an OWL-based air quality ontology extending the Semantic Sensor Network (SSN) ontology is developed to represent pollutants, monitoring stations, AQI categories, regulatory standards, and environmental governance actions. Semantic reasoning is implemented using SWRL rules and validated through SPARQL queries to infer AQI categories, health risks, and recommended mitigation actions. Experimental evaluation using CPCB air quality datasets demonstrates that the proposed framework improves AQI classification reliability and uncertainty handling compared with traditional crisp and Type-1 fuzzy approaches, while enabling explainable semantic reasoning and intelligent decision support for air quality monitoring systems
【3】ICLAD: In-Context Learning for Unified Tabular Anomaly Detection Across Supervision Regimes
标题：ICLAT：跨监管制度统一表格异常检测的上下文学习
链接：https://arxiv.org/abs/2603.19497
备注：33 pages, 17 figures
摘要：表格数据的异常检测通常在三种监督机制下进行研究，包括一类设置，假设可以访问无异常的训练样本，完全无监督的设置未标记和潜在污染的训练数据，以及具有有限异常标签的半监督设置。现有的深度学习方法通常在单一监督机制的假设下训练特定于网络的模型，这限制了它们在异常检测任务中利用共享结构和适应不同监督级别的能力。我们提出了ICLAD，一个上下文学习的基础模型，表格异常检测，概括了两个数据集和监督制度。ICLAD通过综合表格异常检测任务的元学习进行训练，在推理时，模型通过调整训练集来分配异常分数，而无需更新模型权重。在ADBench的57个表格数据集上进行的综合实验表明，该方法在三种监督机制下均达到了最先进的性能，为表格异常检测建立了统一的框架。
摘要：Anomaly detection on tabular data is commonly studied under three supervision regimes, including one-class settings that assume access to anomaly-free training samples, fully unsupervised settings with unlabeled and potentially contaminated training data, and semi-supervised settings with limited anomaly labels. Existing deep learning approaches typically train dataset-specific models under the assumption of a single supervision regime, which limits their ability to leverage shared structures across anomaly detection tasks and to adapt to different supervision levels. We propose ICLAD, an in-context learning foundation model for tabular anomaly detection that generalizes across both datasets and supervision regimes. ICLAD is trained via meta-learning on synthetic tabular anomaly detection tasks, and at inference time, the model assigns anomaly scores by conditioning on the training set without updating model weights. Comprehensive experiments on 57 tabular datasets from ADBench show that our method achieves state-of-the-art performance across three supervision regimes, establishing a unified framework for tabular anomaly detection.
【4】Pseudo-Labeling for Unsupervised Domain Adaptation with Kernel GLMs
标题：基于核GLM的无监督域自适应伪标记
链接：https://arxiv.org/abs/2603.19422
备注：55 pages, 4 figures. Python solvers and experiment scripts are available at: https://github.com/nathanweill/KRGLM
摘要：我们提出了一个原则性框架，用于核广义线性模型（GLM）中协变量移位下的无监督域自适应，包括带岭正则化的核化线性，逻辑和泊松回归。我们的目标是通过利用标记的源数据和未标记的目标数据来最小化目标域中的预测误差，尽管协变量分布存在差异。我们将标记的源数据分为两批：一批用于训练一系列候选模型，另一批用于构建插补模型。该插补模型为目标数据生成伪标签，从而实现稳健的模型选择。我们建立了非渐近的过度风险界限，通过“有效标记样本量”来表征适应性能，明确说明未知的协变量变化。在合成数据集和真实数据集上的实验表明，与仅源基线相比，性能得到了一致的提高。
摘要：We propose a principled framework for unsupervised domain adaptation under covariate shift in kernel Generalized Linear Models (GLMs), encompassing kernelized linear, logistic, and Poisson regression with ridge regularization. Our goal is to minimize prediction error in the target domain by leveraging labeled source data and unlabeled target data, despite differences in covariate distributions. We partition the labeled source data into two batches: one for training a family of candidate models, and the other for building an imputation model. This imputation model generates pseudo-labels for the target data, enabling robust model selection. We establish non-asymptotic excess-risk bounds that characterize adaptation performance through an "effective labeled sample size", explicitly accounting for the unknown covariate shift. Experiments on synthetic and real datasets demonstrate consistent performance gains over source-only baselines.
迁移|Zero/Few/One-Shot|自适应(5篇)
【1】Evaluating Test-Time Adaptation For Facial Expression Recognition Under Natural Cross-Dataset Distribution Shifts
标题：自然跨数据集分布变化下评估面部表情识别的测试时间适应性
链接：https://arxiv.org/abs/2603.19994
备注：Accepted at ICASSP 2026
摘要：深度学习模型经常在自然分布变化下挣扎，这是现实世界部署中的常见挑战。测试时自适应（TTA）通过在没有标记源数据的情况下在推理过程中调整模型来解决这个问题。我们提出了第一次评估的TTA方法FER自然域的变化，广泛使用的FER数据集进行跨数据集实验。这超越了合成腐败，以检查由不同的收集协议，注释标准和人口统计引起的真实世界的变化。结果表明，TTA可以使自然漂移下的FER性能提高11.34%.当目标分布干净时，熵最小化方法（如TENT和SAR）的性能最好。相比之下，T3 A等原型平差方法在更大的分布距离场景下表现出色。最后，当目标分布比源分布更嘈杂时，SHOT等特征对齐方法可以提供最大的增益。我们的跨数据集分析表明，TTA的有效性是由分布的距离和跨域的自然转移的严重程度。
摘要 ：Deep learning models often struggle under natural distribution shifts, a common challenge in real-world deployments. Test-Time Adaptation (TTA) addresses this by adapting models during inference without labeled source data. We present the first evaluation of TTA methods for FER under natural domain shifts, performing cross-dataset experiments with widely used FER datasets. This moves beyond synthetic corruptions to examine real-world shifts caused by differing collection protocols, annotation standards, and demographics. Results show TTA can boost FER performance under natural shifts by up to 11.34\%. Entropy minimization methods such as TENT and SAR perform best when the target distribution is clean. In contrast, prototype adjustment methods like T3A excel under larger distributional distance scenarios. Finally, feature alignment methods such as SHOT deliver the largest gains when the target distribution is noisier than our source. Our cross-dataset analysis shows that TTA effectiveness is governed by the distributional distance and the severity of the natural shift across domains.
【2】ATHENA: Adaptive Test-Time Steering for Improving Count Fidelity in Diffusion Models
标题：ATHENA：自适应测试时引导，用于提高扩散模型中的计数保真度
链接：https://arxiv.org/abs/2603.19676
摘要：文本到图像的扩散模型实现了高的视觉保真度，但令人惊讶的是，当提示指定明确的对象计数时，在数字控制中表现出系统故障。为了解决这一限制，我们引入ATHENA，一个模型不可知的，测试时自适应转向框架，提高对象计数保真度，而无需修改模型架构或需要重新训练。ATHENA在采样期间利用中间表示来估计对象计数，并在去噪过程的早期应用计数感知噪声校正，在结构错误变得难以修改之前控制生成轨迹。我们提出了三个渐进的ATHENA更先进的变种，贸易额外的计算，提高数值精度，从静态的基于方向盘的动态调整计数感知控制。在已建立的基准测试和一个新的视觉和语义复杂的数据集上进行的实验表明，ATHENA始终提高了计数保真度，特别是在较高的目标计数时，同时在多个扩散主干上保持了良好的准确性-运行时权衡。
摘要：Text-to-image diffusion models achieve high visual fidelity but surprisingly exhibit systematic failures in numerical control when prompts specify explicit object counts. To address this limitation, we introduce ATHENA, a model-agnostic, test-time adaptive steering framework that improves object count fidelity without modifying model architectures or requiring retraining. ATHENA leverages intermediate representations during sampling to estimate object counts and applies count-aware noise corrections early in the denoising process, steering the generation trajectory before structural errors become difficult to revise. We present three progressively more advanced variants of ATHENA that trade additional computation for improved numerical accuracy, ranging from static prompt-based steering to dynamically adjusted count-aware control. Experiments on established benchmarks and a new visually and semantically complex dataset show that ATHENA consistently improves count fidelity, particularly at higher target counts, while maintaining favorable accuracy-runtime trade-offs across multiple diffusion backbones.
【3】An Adaptive Machine Learning Framework for Fluid Flow in Dual-Network Porous Media
标题：双网络多孔介质中流体流动的自适应机器学习框架
链接：https://arxiv.org/abs/2603.19561
摘要：天然或工程多孔材料通常具有双重孔隙网络结构，这些结构控制着矿物勘探和从致密页岩中回收碳氢化合物等过程。双孔隙度/渗透率（DPP）数学模型描述了不可压缩流体在两个相互作用的孔隙网络中的流动，网络间存在质量交换。尽管数值方法取得了重大进展，但仍然需要能够实现快速预报、数据同化和可靠逆分析的计算框架。为了解决这个问题，我们提出了一个物理信息的神经网络（PINN）框架的DPP系统的正向和逆向建模。所提出的方法编码的混合形式的控制方程，随着边界条件，直接到损失函数，自适应加权策略，以平衡它们的贡献。该框架的主要特征包括自适应权重调整，动态配置点选择，以及使用共享主干神经架构来有效地捕获双孔网络的耦合行为。它本质上是无网格的，使其非常适合多孔介质典型的复杂几何形状。它准确地捕捉跨分层域的解域中的不连续性，而不会引入在经典有限元公式中常见的寄生振荡。重要的是，该框架非常适合逆分析，在关键物理量（如DPP模型中的传质系数）难以直接测量的情况下，能够进行稳健的参数识别。此外，一个系统的收敛性分析，严格评估的稳定性，准确性和可靠性的方法。通过一系列有代表性的数值实验证明了该方法的有效性和计算优势。
摘要：Porous materials -- natural or engineered -- often exhibit dual pore-network structures that govern processes such as mineral exploration and hydrocarbon recovery from tight shales. Double porosity/permeability (DPP) mathematical models describe incompressible fluid flow through two interacting pore networks with inter-network mass exchange. Despite significant advances in numerical methods, there remains a need for computational frameworks that enable rapid forecasting, data assimilation, and reliable inverse analysis. To address this, we present a physics-informed neural network (PINN) framework for forward and inverse modeling of DPP systems. The proposed approach encodes the governing equations in mixed form, along with boundary conditions, directly into the loss function, with adaptive weighting strategies to balance their contributions. Key features of the framework include adaptive weight tuning, dynamic collocation point selection, and the use of shared trunk neural architectures to efficiently capture the coupled behavior of the dual pore networks. It is inherently mesh-free, making it well-suited for complex geometries typical of porous media. It accurately captures discontinuities in solution fields across layered domains without introducing spurious oscillations commonly observed in classical finite element formulations. Importantly, the framework is well-suited for inverse analysis, enabling robust parameter identification in scenarios where key physical quantities -- such as the mass transfer coefficient in DPP models -- are difficult to measure directly. In addition, a systematic convergence analysis is provided to rigorously assess the stability, accuracy, and reliability of the method. The effectiveness and computational advantages of the approach are demonstrated through a series of representative numerical experiments.
【4】DPxFin: Adaptive Differential Privacy for Anti-Money Laundering Detection via Reputation-Weighted Federated Learning
标题：DPxFin：通过声誉加权联邦学习进行反洗钱检测的自适应差异隐私
链接：https://arxiv.org/abs/2603.19314
备注：Accepted at AI FOR FINANCIAL FRAUD DETECTION & PREVENTION AT ACM ICAIF-25
摘要：在现代金融体系中，打击洗钱是一项关键挑战，数据隐私问题和日益复杂的欺诈交易模式使其变得更加复杂。虽然联邦学习（FL）是一种很有前途的解决问题的方法，因为它允许机构在不共享数据的情况下训练他们的模型，但它有容易泄露隐私的缺点，特别是在财务数据等表格数据形式中。为了解决这个问题，我们提出了DPxFin，一种新的联邦框架，集成了声誉引导的自适应差分隐私。我们的方法通过评估本地训练模型和全局模型之间的对齐来计算客户端声誉。基于此声誉，我们动态分配差分隐私噪声客户端更新，提高隐私，同时保持整体模型的效用。声誉较高的客户会收到较低的噪音，以放大他们值得信赖的贡献，而声誉较低的客户会收到较强的噪音，以减轻风险。我们使用多层感知器（MLP）在IID和非IID设置下在反洗钱（AML）数据集上验证DPxFin。实验分析表明，我们的方法比传统FL和固定噪声差分隐私（DP）基线在准确性和隐私性之间具有更理想的权衡，即使在适度的规模上，性能改进也是一致的。此外，DPxFin确实能够抵御表格数据泄漏攻击，证明了其在现实金融条件下的有效性。
摘要 ：In the modern financial system, combating money laundering is a critical challenge complicated by data privacy concerns and increasingly complex fraud transaction patterns. Although federated learning (FL) is a promising problem-solving approach as it allows institutions to train their models without sharing their data, it has the drawback of being prone to privacy leakage, specifically in tabular data forms like financial data. To address this, we propose DPxFin, a novel federated framework that integrates reputation-guided adaptive differential privacy. Our approach computes client reputation by evaluating the alignment between locally trained models and the global model. Based on this reputation, we dynamically assign differential privacy noise to client updates, enhancing privacy while maintaining overall model utility. Clients with higher reputations receive lower noise to amplify their trustworthy contributions, while low-reputation clients are allocated stronger noise to mitigate risk. We validate DPxFin on the Anti-Money Laundering (AML) dataset under both IID and non-IID settings using Multi Layer Perceptron (MLP). Experimental analysis established that our approach has a more desirable trade-off between accuracy and privacy than those of traditional FL and fixed-noise Differential Privacy (DP) baselines, where performance improvements were consistent, even though on a modest scale. Moreover, DPxFin does withstand tabular data leakage attacks, proving its effectiveness under real-world financial conditions.
【5】Minimax and Adaptive Covariance Matrix Estimation under Differential Privacy
标题：差异隐私下的极小极大和自适应协方差矩阵估计
链接：https://arxiv.org/abs/2603.19703
摘要：协方差矩阵在高维数据的分析中起着重要的作用。本文研究了差分隐私约束下高维可带协方差矩阵的极小极大估计和自适应估计。我们提出了一种新的差分私有块三对角估计，实现最小最优收敛速度下的运营商范数和Frobenius范数。与非隐私设置相反，隐私引起的错误表现出对环境维度的多项式依赖性，揭示了大量的额外隐私成本。 为了建立最优性，我们开发了一个新的差分私有van Trees不等式，并构造了精心设计的先验分布，以获得匹配的minimax下界。所提出的私有van Trees不等式更广泛地适用于一般的私有估计问题，并且具有独立的意义。我们进一步介绍了一种自适应估计，达到最佳的速率对数因子没有先验知识的衰减参数，基于一种新的分层三对角方法。数值实验证实了理论结果，并说明了基本的隐私准确性权衡。
摘要：The covariance matrix plays a fundamental role in the analysis of high-dimensional data. This paper studies minimax and adaptive estimation of high-dimensional bandable covariance matrices under differential privacy constraints. We propose a novel differentially private blockwise tridiagonal estimator that achieves minimax-optimal convergence rates under both the operator norm and the Frobenius norm. In contrast to the non-private setting, the privacy-induced error exhibits a polynomial dependence on the ambient dimension, revealing a substantial additional cost of privacy. To establish optimality, we develop a new differentially private van Trees inequality and construct carefully designed prior distributions to obtain matching minimax lower bounds. The proposed private van Trees inequality applies more broadly to general private estimation problems and is of independent interest. We further introduce an adaptive estimator that attains the optimal rate up to a logarithmic factor without prior knowledge of the decay parameter, based on a novel hierarchical tridiagonal approach. Numerical experiments corroborate the theoretical results and illustrate the fundamental privacy-accuracy trade-off.
强化学习(5篇)
【1】Fine-tuning Timeseries Predictors Using Reinforcement Learning
标题：使用强化学习微调时间序列预测器
链接：https://arxiv.org/abs/2603.20063
摘要：本章介绍了用于微调金融预测的三种主要强化学习算法。我们提出了一个明确的实施计划，用于将强化学习任务的损失反向传播到使用监督学习训练的模型，并比较微调前后的性能。我们发现微调后的性能有所提高，并将学习属性转移到模型中，表明微调的好处。我们还强调了调整过程和实证结果，为未来的实施从业者。
摘要：This chapter presents three major reinforcement learning algorithms used for fine-tuning financial forecasters. We propose a clear implementation plan for backpropagating the loss of a reinforcement learning task to a model trained using supervised learning, and compare the performance before and after the fine-tuning. We find an increase in performance after fine-tuning, and transfer learning properties to the models, indicating the benefits of fine-tuning. We also highlight the tuning process and empirical results for future implementation by practitioners.
【2】What If Consensus Lies? Selective-Complementary Reinforcement Learning at Test Time
标题：如果共识谎言怎么办？测试时的选择性补充强化学习
链接：https://arxiv.org/abs/2603.19880
备注：14 pages, 5 figures
摘要：测试时强化学习（TTRL）使大型语言模型（LLM）能够通过从多数投票共识中获得伪奖励来增强对未标记测试流的推理能力。然而，现有的TTRL方法完全依赖于积极的伪标记策略。在答案分布高度分散的挑战性场景下，这种依赖变得脆弱，导致共识薄弱，无意中强化了作为监督信号的不正确轨迹。在本文中，我们提出了SCRL（选择互补强化学习），这是一个强大的测试时强化学习框架，可以有效地减轻标签噪声放大。SCRL开发了Selective Positive Pseudo-Labeling，它强制执行严格的共识标准来过滤不可靠的多数。作为补充，SCRL引入了TTRL中的第一个负监督机制熵门负伪标记，以基于生成不确定性可靠地修剪不正确的轨迹。在多个推理基准上进行的大量实验表明，SCRL在基线上实现了实质性的改进，同时在受限的推出预算下保持了鲁棒的泛化和训练稳定性。我们的代码可在https://github.com/Jasper-Yan/SCRL上获得。
摘要：Test-Time Reinforcement Learning (TTRL) enables Large Language Models (LLMs) to enhance reasoning capabilities on unlabeled test streams by deriving pseudo-rewards from majority voting consensus. However, existing TTRL methods rely exclusively on positive pseudo-labeling strategies. Such reliance becomes vulnerable under challenging scenarios where answer distributions are highly dispersed, resulting in weak consensus that inadvertently reinforces incorrect trajectories as supervision signals. In this paper, we propose SCRL (Selective-Complementary Reinforcement Learning), a robust test-time reinforcement learning framework that effectively mitigates label noise amplification. SCRL develops Selective Positive Pseudo-Labeling, which enforces strict consensus criteria to filter unreliable majorities. Complementarily, SCRL introduces Entropy-Gated Negative Pseudo-Labeling, the first negative supervision mechanism in TTRL, to reliably prune incorrect trajectories based on generation uncertainty. Extensive experiments on multiple reasoning benchmarks demonstrate that SCRL achieves substantial improvements over baselines, while maintaining robust generalization and training stability under constrained rollout budgets. Our code is available at https://github.com/Jasper-Yan/SCRL.
【3】DeepStock: Reinforcement Learning with Policy Regularizations for Inventory Management
标题：DeepStock：库存管理的强化学习和政策监管
链接：https://arxiv.org/abs/2603.19621
摘要：深度强化学习（DRL）提供了一种通用方法，用于训练可以利用大数据和计算的库存策略。然而，DRL的现成实现取得了混合的成功，通常受到训练期间使用的超参数的高敏感性的困扰。在本文中，我们表明，通过施加政策正则化，接地在经典的库存概念，如“基本库存”，我们可以显着加快超参数调整，提高最终性能的几个DRL方法。我们报告了阿里巴巴电子商务平台天猫上100%部署DRL的详细信息。我们还包括广泛的合成实验，这表明，政策规范化重塑叙述什么是最好的DRL方法的库存管理。
摘要 ：Deep Reinforcement Learning (DRL) provides a general-purpose methodology for training inventory policies that can leverage big data and compute. However, off-the-shelf implementations of DRL have seen mixed success, often plagued by high sensitivity to the hyperparameters used during training. In this paper, we show that by imposing policy regularizations, grounded in classical inventory concepts such as "Base Stock", we can significantly accelerate hyperparameter tuning and improve the final performance of several DRL methods. We report details from a 100% deployment of DRL with policy regularizations on Alibaba's e-commerce platform, Tmall. We also include extensive synthetic experiments, which show that policy regularizations reshape the narrative on what is the best DRL method for inventory management.
【4】PA2D-MORL: Pareto Ascent Directional Decomposition based Multi-Objective Reinforcement Learning
标题：PA 2D-MORL：基于帕累托上升方向分解的多目标强化学习
链接：https://arxiv.org/abs/2603.19579
备注：AAAI 2024
摘要：多目标强化学习（MORL）为解决目标冲突的决策问题提供了一种有效的方法。然而，实现高质量的近似帕累托策略集仍然具有挑战性，特别是在连续或高维状态动作空间的复杂任务。本文提出了基于Pareto上升方向分解的多目标强化学习（PA 2D-MORL）方法，该方法构造了一个有效的多目标问题分解和策略改进方案，从而得到了Pareto策略集的一个更优近似。该方法利用Pareto上升方向选择标量化权值，计算多目标策略梯度，确定策略优化方向，实现各目标的联合改进。同时，多个政策的进化框架下，有选择地优化，从不同的方向接近帕累托前沿。此外，一个帕累托自适应微调的方法，以提高密度和传播的帕累托前沿近似。各种多目标机器人控制任务的实验表明，所提出的方法明显优于当前国家的最先进的算法的质量和稳定性的结果。
摘要：Multi-objective reinforcement learning (MORL) provides an effective solution for decision-making problems involving conflicting objectives. However, achieving high-quality approximations to the Pareto policy set remains challenging, especially in complex tasks with continuous or high-dimensional state-action space. In this paper, we propose the Pareto Ascent Directional Decomposition based Multi-Objective Reinforcement Learning (PA2D-MORL) method, which constructs an efficient scheme for multi-objective problem decomposition and policy improvement, leading to a superior approximation of Pareto policy set. The proposed method leverages Pareto ascent direction to select the scalarization weights and computes the multi-objective policy gradient, which determines the policy optimization direction and ensures joint improvement on all objectives. Meanwhile, multiple policies are selectively optimized under an evolutionary framework to approximate the Pareto frontier from different directions. Additionally, a Pareto adaptive fine-tuning approach is applied to enhance the density and spread of the Pareto frontier approximation. Experiments on various multi-objective robot control tasks show that the proposed method clearly outperforms the current state-of-the-art algorithm in terms of both quality and stability of the outcomes.
【5】Optimizing Resource-Constrained Non-Pharmaceutical Interventions for Multi-Cluster Outbreak Control Using Hierarchical Reinforcement Learning
标题：使用分层强化学习优化资源限制的非药物干预措施以控制多集群疫情
链接：https://arxiv.org/abs/2603.19397
摘要：诊断检测和检疫等非药物干预措施对于控制传染病暴发至关重要，但往往受到资源有限的限制，特别是在暴发的早期阶段。在现实世界的公共卫生环境中，必须在多个爆发集群之间分配资源，这些集群异步出现，规模和风险各不相同，并争夺共享的资源预算。在这里，一个集群对应于由单个感染的索引病例产生的一组密切接触者。因此，必须在不确定性和不同需求的情况下做出决策，同时尊重业务限制。我们将这个问题描述为一个受约束的不安分的多臂强盗，并提出了一个分层强化学习框架。一个全局控制器学习一个连续的行动成本乘数，调整全球资源需求，而一个广义的本地政策估计分配资源给每个集群内的个人的边际值。我们评估所提出的框架在一个现实的基于代理的模拟器的SARS-CoV-2的动态到达集群。在广泛的系统规模和测试预算范围内，我们的方法始终优于RMAB启发式和启发式基线，将爆发控制有效性提高了20%-30%。多达40个并发活动的集群上的实验进一步证明，分层框架是高度可扩展的，并使更快的决策比RMAB启发的方法。
摘要：Non-pharmaceutical interventions (NPIs), such as diagnostic testing and quarantine, are crucial for controlling infectious disease outbreaks but are often constrained by limited resources, particularly in early outbreak stages. In real-world public health settings, resources must be allocated across multiple outbreak clusters that emerge asynchronously, vary in size and risk, and compete for a shared resource budget. Here, a cluster corresponds to a group of close contacts generated by a single infected index case. Thus, decisions must be made under uncertainty and heterogeneous demands, while respecting operational constraints. We formulate this problem as a constrained restless multi-armed bandit and propose a hierarchical reinforcement learning framework. A global controller learns a continuous action cost multiplier that adjusts global resource demand, while a generalized local policy estimates the marginal value of allocating resources to individuals within each cluster. We evaluate the proposed framework in a realistic agent-based simulator of SARS-CoV-2 with dynamically arriving clusters. Across a wide range of system scales and testing budgets, our method consistently outperforms RMAB-inspired and heuristic baselines, improving outbreak control effectiveness by 20%-30%. Experiments on up to 40 concurrently active clusters further demonstrate that the hierarchical framework is highly scalable and enables faster decision-making than the RMAB-inspired method.
医学相关(5篇)
【1】ReXInTheWild: A Unified Benchmark for Medical Photograph Understanding
标题：ReXInTheWild：医学照片理解的统一基准
链接：https://arxiv.org/abs/2603.19517
备注：11 pages, 4 figures
摘要：用普通相机拍摄的日常照片已经广泛用于远程医疗和其他在线健康对话，但没有全面的基准评估视觉语言模型是否可以解释其医疗内容。分析这些图像需要细粒度的自然图像理解和特定领域的医学推理，这一组合对通用和专用模型都提出了挑战。我们介绍ReXInTheWild，这是一个由955个临床医生验证的多项选择题组成的基准，涵盖了来自生物医学文献的484张照片中的7个临床主题。在ReXInTheWild上进行评估时，领先的多模态大型语言模型显示出显著的性能差异：Gemini-3达到了78%的准确率，其次是Claude Opus 4.5（72%）和GPT-5（68%），而医学专家模型MedGemma仅达到37%。系统的错误分析还揭示了四类常见的错误，从低层次的几何错误到高层次的推理失败，需要不同的缓解策略。ReXInTheWild在自然图像理解和医学推理的交叉点提供了一个具有挑战性的、基于临床的基准。该数据集可在HuggingFace上使用。
摘要：Everyday photographs taken with ordinary cameras are already widely used in telemedicine and other online health conversations, yet no comprehensive benchmark evaluates whether vision-language models can interpret their medical content. Analyzing these images requires both fine-grained natural image understanding and domain-specific medical reasoning, a combination that challenges both general-purpose and specialized models. We introduce ReXInTheWild, a benchmark of 955 clinician-verified multiple-choice questions spanning seven clinical topics across 484 photographs sourced from the biomedical literature. When evaluated on ReXInTheWild, leading multimodal large language models show substantial performance variation: Gemini-3 achieves 78% accuracy, followed by Claude Opus 4.5 (72%) and GPT-5 (68%), while the medical specialist model MedGemma reaches only 37%. A systematic error analysis also reveals four categories of common errors, ranging from low-level geometric errors to high-level reasoning failures and requiring different mitigation strategies. ReXInTheWild provides a challenging, clinically grounded benchmark at the intersection of natural image understanding and medical reasoning. The dataset is available on HuggingFace.
【2】Parameter-Efficient Token Embedding Editing for Clinical Class-Level Unlearning
标题：用于临床班级级取消学习的参数高效令牌嵌入编辑
链接：https://arxiv.org/abs/2603.19302
备注：10 pages
摘要：机器非学习对于临床语言模型越来越重要，隐私法规和机构政策可能需要从部署的系统中删除敏感信息，而无需从头开始重新训练。在实践中，删除请求必须在目标信息的有效遗忘与模型效用的保留和最小参数修改之间取得平衡。我们介绍了稀疏令牌嵌入非学习（STEU），这是一种用于行为类级非学习的参数高效方法，它只更新PMI选择的令牌嵌入以及一个小的分类器头，同时保持所有编码器层冻结。在使用BioClinicalBERT、BERT-base和DistilBERT对MIMIC-IV、MIMIC-III和eICU进行的实验中，STEU始终抑制目标类，同时在很大程度上保留了任务性能。在主要MIMIC-IV设置中，STEU在仅修改0.19%的模型参数后实现了接近完全的遗忘（遗忘F1 = 0.0004），同时保持了竞争性保留效用（保留平均F1 = 0.4766）。这些结果表明，可以通过稀疏嵌入编辑实现有针对性的行为遗忘，而无需修改更深的编码器表示。
摘要：Machine unlearning is increasingly important for clinical language models, where privacy regulations and institutional policies may require removing sensitive information from deployed systems without retraining from scratch. In practice, deletion requests must balance effective forgetting of targeted information with preservation of model utility and minimal parameter modification. We introduce Sparse Token Embedding Unlearning (STEU), a parameter-efficient method for behavioral class-level unlearning that updates only PMI-selected token embeddings together with a small classifier head while keeping all encoder layers frozen. Across experiments on MIMIC-IV, MIMIC-III, and eICU using BioClinicalBERT, BERT-base, and DistilBERT, STEU consistently suppresses the target class while largely preserving retained task performance. In the primary MIMIC-IV setting, STEU achieves near-complete forgetting (forget F1 = 0.0004) while maintaining competitive retained utility (retain avg F1 = 0.4766) after modifying only 0.19\% of model parameters. These results suggest that targeted behavioral unlearning can be achieved through sparse embedding edits without modifying deeper encoder representations.
【3】PRIME-CVD: A Parametrically Rendered Informatics Medical Environment for Education in Cardiovascular Risk Modelling
标题：PRME-VD：用于心血管风险建模教育的参数化信息学医疗环境
链接：https://arxiv.org/abs/2603.19299
摘要：近年来，医学信息学和机器学习的进步由于开放访问的基准数据集的可用性而加速。然而，由于隐私、治理和重新识别风险，患者级电子病历（EMR）数据很少可用于教学或方法开发。这限制了心血管风险建模的可重复性，透明度和实践培训。在这里，我们介绍PRIME-CVD，一个参数化呈现的信息学医疗环境明确设计的医学教育。PRIME-CVD包括两个可公开访问的合成数据资产，代表了接受心血管疾病一级预防的50，000名成年人。这些数据集完全由用户指定的因果有向非循环图生成，该图使用公开的澳大利亚人口统计数据和已发表的流行病学效应估计进行参数化，而不是来自患者水平的EMR数据或训练的生成模型。数据资产1提供了一个干净的，分析就绪的队列，适合探索性分析，分层和生存建模，而数据资产2将同一队列重新构建为具有现实结构和词汇异质性的关系型EMR风格数据库。总之，这些资产可以在不暴露敏感信息的情况下指导数据清理、协调、因果推理和政策相关风险建模。由于所有个人和事件都是从头生成的，PRIME-CVD保留了现实的亚组不平衡和风险梯度，同时确保可以忽略的披露风险。PRIME-CVD是在知识共享署名4.0许可下发布的，以支持可重复的研究和可扩展的医学教育。
摘要：In recent years, progress in medical informatics and machine learning has been accelerated by the availability of openly accessible benchmark datasets. However, patient-level electronic medical record (EMR) data are rarely available for teaching or methodological development due to privacy, governance, and re-identification risks. This has limited reproducibility, transparency, and hands-on training in cardiovascular risk modelling. Here we introduce PRIME-CVD, a parametrically rendered informatics medical environment designed explicitly for medical education. PRIME-CVD comprises two openly accessible synthetic data assets representing a cohort of 50,000 adults undergoing primary prevention for cardiovascular disease. The datasets are generated entirely from a user-specified causal directed acyclic graph parameterised using publicly available Australian population statistics and published epidemiologic effect estimates, rather than from patient-level EMR data or trained generative models. Data Asset 1 provides a clean, analysis-ready cohort suitable for exploratory analysis, stratification, and survival modelling, while Data Asset 2 restructures the same cohort into a relational, EMR-style database with realistic structural and lexical heterogeneity. Together, these assets enable instruction in data cleaning, harmonisation, causal reasoning, and policy-relevant risk modelling without exposing sensitive information. Because all individuals and events are generated de novo, PRIME-CVD preserves realistic subgroup imbalance and risk gradients while ensuring negligible disclosure risk. PRIME-CVD is released under a Creative Commons Attribution 4.0 licence to support reproducible research and scalable medical education.
【4】TuLaBM: Tumor-Biased Latent Bridge Matching for Contrast-Enhanced MRI Synthesis
标题：TuLaBM：用于对比增强MRI合成的肿瘤偏置潜在桥匹配
链接：https://arxiv.org/abs/2603.19386
摘要：对比增强磁共振成像（CE-MRI）在脑肿瘤评估中起着至关重要的作用;然而，其采集需要钆基造影剂（GBCA），这增加了成本并引起了安全性问题。因此，从非对比MRI（NC-MRI）合成CE-MRI已成为一种有前途的替代方案。早期基于生成对抗网络（GAN）的方法遭受不稳定性和模式崩溃，而扩散模型尽管具有令人印象深刻的合成质量，但仍然在计算上昂贵，并且通常无法忠实地再现关键的肿瘤对比模式。为了解决这些局限性，我们提出了肿瘤偏置潜在桥匹配（TuLaBM），它将NC到CE MRI翻译公式化为学习潜在空间中源和目标分布之间的布朗桥传输，从而实现有效的训练和推理。为了提高肿瘤区域的保真度，我们引入了一种肿瘤偏向注意机制（TuBAM），该机制在桥进化过程中放大了肿瘤相关的潜在特征，同时还引入了一种边界感知损失，该损失限制了肿瘤界面，以提高边缘清晰度。虽然桥匹配已被探索用于像素空间中的医学图像翻译，但我们的潜在公式大大降低了计算成本和推理时间。在BraTS 2023-GLI（BraSyn）和Cleveland Clinic（内部）肝脏MRI数据集上的实验表明，TuLaBM在整个图像和肿瘤区域指标上始终优于最先进的基线，有效地推广到zero-shot和微调设置中的不可见肝脏MRI数据，并实现了每张图像0.097秒以下的推理时间。
摘要：Contrast-enhanced magnetic resonance imaging (CE-MRI) plays a crucial role in brain tumor assessment; however, its acquisition requires gadolinium-based contrast agents (GBCAs), which increase costs and raise safety concerns. Consequently, synthesizing CE-MRI from non-contrast MRI (NC-MRI) has emerged as a promising alternative. Early Generative Adversarial Network (GAN)-based approaches suffered from instability and mode collapse, while diffusion models, despite impressive synthesis quality, remain computationally expensive and often fail to faithfully reproduce critical tumor contrast patterns. To address these limitations, we propose Tumor-Biased Latent Bridge Matching (TuLaBM), which formulates NC-to-CE MRI translation as Brownian bridge transport between source and target distributions in a learned latent space, enabling efficient training and inference. To enhance tumor-region fidelity, we introduce a Tumor-Biased Attention Mechanism (TuBAM) that amplifies tumor-relevant latent features during bridge evolution, along with a boundary-aware loss that constrains tumor interfaces to improve margin sharpness. While bridge matching has been explored for medical image translation in pixel space, our latent formulation substantially reduces computational cost and inference time. Experiments on BraTS2023-GLI (BraSyn) and Cleveland Clinic (in-house) liver MRI dataset show that TuLaBM consistently outperforms state-of-the-art baselines on both whole-image and tumor-region metrics, generalizes effectively to unseen liver MRI data in zero-shot and fine-tuned settings, and achieves inference times under 0.097 seconds per image.
【5】Mathematical Modeling of Cancer-Bacterial Therapy: Analysis and Numerical Simulation via Physics-Informed Neural Networks
标题：癌症细菌治疗的数学建模：通过物理信息神经网络进行分析和数值模拟
链接：https://arxiv.org/abs/2603.19326
摘要：细菌癌症治疗利用厌氧菌靶向缺氧肿瘤区域的能力，但肿瘤生长、细菌定植、氧水平、免疫抑制细胞因子和细菌通讯之间的相互作用仍然难以量化。我们提出了一个二维组织区域中五个耦合的非线性反应扩散方程的数学模型。我们证明了模型的全局适定性，并确定了其稳定性分析的稳定状态。此外，物理信息神经网络（PINN）在没有网格和不需要大量数据的情况下解决了系统。它通过结合残差稳定性和Sobolev逼近误差界来提供收敛保证。这导致总错误率为O（n^-2 ln^4（n）+ N^-1/2），这取决于网络宽度n和配置点的数量N。我们进行了几个数值实验，包括预测肿瘤对治疗的反应。我们亦对若干参数进行敏感度分析。结果表明，长期治疗效果可能需要维持肿瘤中的缺氧区域，或者使用耐受氧气更好的细菌，这可能是长期控制肿瘤所必需的。
摘要：Bacterial cancer therapy exploits anaerobic bacteria's ability to target hypoxia tumor regions, yet the interactions among tumor growth, bacterial colonization, oxygen levels, immunosuppressive cytokines, and bacterial communication remain poorly quantified. We present a mathematical model of five coupled nonlinear reaction-diffusion equations in a two-dimensional tissue domain. We proved the global well-posedness of the model and identified its steady states to analyze stability. Furthermore, a physics-informed neural network (PINN) solves the system without a mesh and without requiring extensive data. It provides convergence guarantees by combining residual stability and Sobolev approximation error bounds. This results in an overall error rate of O(n^-2 ln^4(n) + N^-1/2), which depends on the network width n and the number of collocation points N. We conducted several numerical experiments, including predicting the tumor's response to therapy. We also performed a sensitivity analysis of certain parameters. The results suggest that long-term therapeutic efficacy may require the maintenance of hypoxia regions in the tumor, or using bacteria that tolerate oxygen better, may be necessary for long-lasting tumor control.
蒸馏|知识提取(1篇)
【1】Beyond Single Tokens: Distilling Discrete Diffusion Models via Discrete MMD
标题：超越单一代币：通过离散MMD提取离散扩散模型
链接：https://arxiv.org/abs/2603.20155
摘要：目前，很难提取离散扩散模型。相比之下，连续扩散文献有许多蒸馏方法，可以将采样步骤减少到少数。 我们的方法，离散矩匹配蒸馏（D-MMD），利用了在连续领域非常成功的想法。虽然以前的离散蒸馏方法崩溃，D-MMD保持高质量和多样性（给定足够的采样步骤）。这在文本和图像数据集上都得到了证明。此外，新提炼的发电机可以超越他们的老师。
摘要：It is currently difficult to distill discrete diffusion models. In contrast, continuous diffusion literature has many distillation approaches methods that can reduce sampling steps to a handful. Our method, Discrete Moment Matching Distillation (D-MMD), leverages ideas that have been highly successful in the continuous domain. Whereas previous discrete distillation methods collapse, D-MMD maintains high quality and diversity (given sufficient sampling steps). This is demonstrated on both text and image datasets. Moreover, the newly distilled generators can outperform their teachers.
聚类(2篇)
【1】MOSAIC: Modular Opinion Summarization using Aspect Identification and Clustering
标题：MOSAIC：使用方面识别和聚类的模块化意见摘要
链接：https://arxiv.org/abs/2603.19277
摘要：评论是旅行者如何评估在线市场上产品的核心，但现有的总结研究往往强调端到端的质量，而忽视了基准的可靠性和粒度洞察的实际效用。为了解决这个问题，我们提出了MOSAIC，一个可扩展的，模块化的框架，专为工业部署，分解总结成可解释的组件，包括主题发现，结构化的意见提取，接地摘要生成。我们通过在实时产品页面上进行在线A/B测试来验证我们的方法的实际影响，表明即使在全面总结部署之前，呈现中间输出也可以改善客户体验并提供可衡量的价值。我们进一步进行了广泛的离线实验，以证明MOSAIC实现了优越的方面覆盖率和忠实性相比，强大的基线摘要。至关重要的是，我们引入意见聚类作为一个系统级组件，并表明它显着提高忠诚度，特别是在嘈杂和冗余的条件下，典型的用户评论。最后，我们确定了标准SPACE数据集的可靠性限制，并发布了一个新的开源旅游体验数据集（TRECS），以实现更强大的评估。
摘要：Reviews are central to how travelers evaluate products on online marketplaces, yet existing summarization research often emphasizes end-to-end quality while overlooking benchmark reliability and the practical utility of granular insights. To address this, we propose MOSAIC, a scalable, modular framework designed for industrial deployment that decomposes summarization into interpretable components, including theme discovery, structured opinion extraction, and grounded summary generation. We validate the practical impact of our approach through online A/B tests on live product pages, showing that surfacing intermediate outputs improves customer experience and delivers measurable value even prior to full summarization deployment. We further conduct extensive offline experiments to demonstrate that MOSAIC achieves superior aspect coverage and faithfulness compared to strong baselines for summarization. Crucially, we introduce opinion clustering as a system-level component and show that it significantly enhances faithfulness, particularly under the noisy and redundant conditions typical of user reviews. Finally, we identify reliability limitations in the standard SPACE dataset and release a new open-source tour experience dataset (TRECS) to enable more robust evaluation.
【2】Explainable cluster analysis: a bagging approach
标题：可解释的集群分析：袋装方法
链接：https://arxiv.org/abs/2603.19840
摘要：聚类方法的一个主要限制是它们缺乏可解释性：方法很少提供对哪些特征驱动相似观测分组的洞察。为了解决这一限制，我们提出了一个基于集合的聚类框架，集成装袋和功能辍学生成功能的重要性分数，在类似的功能重要性机制，在监督随机森林。通过利用多个自举恢复方案并聚合所得到的分区，该方法提高了聚类定义的稳定性和鲁棒性，特别是在小样本或噪声设置中。特征重要性通过信息理论方法进行评估：在每一步中，计算每个特征和估计的聚类标签之间的互信息，并通过聚类有效性的度量进行加权，以强调格式良好的分区，然后聚合为最终得分。该方法输出的共识分区和相应的功能的重要性措施，使聚类结构和变量相关性的统一解释。它的有效性在多个模拟和真实世界的数据集上得到了证明。
摘要：A major limitation of clustering approaches is their lack of explainability: methods rarely provide insight into which features drive the grouping of similar observations. To address this limitation, we propose an ensemble-based clustering framework that integrates bagging and feature dropout to generate feature importance scores, in analogy with feature importance mechanisms in supervised random forests. By leveraging multiple bootstrap resampling schemes and aggregating the resulting partitions, the method improves stability and robustness of the cluster definition, particularly in small-sample or noisy settings. Feature importance is assessed through an information-theoretic approach: at each step, the mutual information between each feature and the estimated cluster labels is computed and weighted by a measure of clustering validity to emphasize well-formed partitions, before being aggregated into a final score. The method outputs both a consensus partition and a corresponding measure of feature importance, enabling a unified interpretation of clustering structure and variable relevance. Its effectiveness is demonstrated on multiple simulated and real-world datasets.
点云|SLAM|雷达|激光|深度RGBD相关(1篇)
【1】Layered Quantum Architecture Search for 3D Point Cloud Classification
标题：基于分层量子结构的三维点云分类算法
链接：https://arxiv.org/abs/2603.20024
摘要：我们引入分层量子架构搜索（分层QAS），这是一种受经典网络态射启发的策略，该策略通过逐步增长和调整来设计参数化量子电路（PQC）架构。PQC以相对较少的参数提供了强大的表现力，但它们缺乏标准的架构层（例如，卷积，注意力），其对给定学习任务的归纳偏差进行编码。为了评估我们的方法的有效性，我们专注于三维点云分类作为一个具有挑战性的，但高度结构化的问题。鉴于这项任务的先前工作仅使用PQC作为经典分类器的特征提取器，我们的方法使用PQC作为分类模型的主要构建块。模拟表明，我们的分层QAS缓解了贫瘠的高原，优于量子适应的本地和进化QAS基线，并在ModelNet数据集上实现了基于PQC的方法中最先进的结果。
摘要：We introduce layered Quantum Architecture Search (layered-QAS), a strategy inspired by classical network morphism that designs Parametrised Quantum Circuit (PQC) architectures by progressively growing and adapting them. PQCs offer strong expressiveness with relatively few parameters, yet they lack standard architectural layers (e.g., convolution, attention) that encode inductive biases for a given learning task. To assess the effectiveness of our method, we focus on 3D point cloud classification as a challenging yet highly structured problem. Whereas prior work on this task has used PQCs only as feature extractors for classical classifiers, our approach uses the PQC as the main building block of the classification model. Simulations show that our layered-QAS mitigates barren plateau, outperforms quantum-adapted local and evolutionary QAS baselines, and achieves state-of-the-art results among PQC-based methods on the ModelNet dataset.
联邦学习|隐私保护|加密(3篇)
【1】FedRG: Unleashing the Representation Geometry for Federated Learning with Noisy Clients
标题：FedRG：释放具有噪音客户端的联邦学习的表示几何
链接：https://arxiv.org/abs/2603.19722
备注：conference
摘要：联邦学习（FL）的性能下降，由于不可避免地存在嘈杂的注释在分布式场景。现有的方法已经通过利用损失值在区分噪声样本与数据集以进行标签校正方面取得了进展。然而，噪声样本识别依赖于标量损失缺乏可靠性的FL异构场景下。在本文中，我们重新思考这一范式从一个代表性的角度来看，并提出\方法~（\textbf{Fed}erated under \textbf{R}epresentation \textbf{G}emometry），它遵循\textbf{原则的“表征几何优先级”}识别噪声标签。首先，\method~通过使用自我监督来创建与标签无关的球形表示。然后，它迭代地拟合一个球形von Mises-Fisher（vMF）混合模型，使用先前确定的干净样本来捕获语义聚类。该几何证据与语义标签软映射机制相结合，以推导出无标签和注释标签条件特征空间之间的分布发散，其鲁棒地识别噪声样本并使用新分离的干净数据集更新vMF混合模型。最后，我们采用一个额外的个性化噪声吸收矩阵的噪声标签，以实现强大的优化。大量的实验结果表明，\方法~显着优于国家的最先进的方法FL与数据异构性下，不同的嘈杂的客户端场景。
摘要：Federated learning (FL) suffers from performance degradation due to the inevitable presence of noisy annotations in distributed scenarios. Existing approaches have advanced in distinguishing noisy samples from the dataset for label correction by leveraging loss values. However, noisy samples recognition relying on scalar loss lacks reliability for FL under heterogeneous scenarios. In this paper, we rethink this paradigm from a representation perspective and propose \method~(\textbf{Fed}erated under \textbf{R}epresentation \textbf{G}emometry), which follows \textbf{the principle of ``representation geometry priority''} to recognize noisy labels. Firstly, \method~creates label-agnostic spherical representations by using self-supervision. It then iteratively fits a spherical von Mises-Fisher (vMF) mixture model to this geometry using previously identified clean samples to capture semantic clusters. This geometric evidence is integrated with a semantic-label soft mapping mechanism to derive a distribution divergence between the label-free and annotated label-conditioned feature space, which robustly identifies noisy samples and updates the vMF mixture model with the newly separated clean dataset. Lastly, we employ an additional personalized noise absorption matrix on noisy labels to achieve robust optimization. Extensive experimental results demonstrate that \method~significantly outperforms state-of-the-art methods for FL with data heterogeneity under diverse noisy clients scenarios.
【2】On Performance Guarantees for Federated Learning with Personalized Constraints
标题：具有个性化约束的联邦学习的性能保证
链接：https://arxiv.org/abs/2603.19617
摘要：联邦学习（FL）已经成为一个跨多个代理的分布式学习的通信效率的算法框架。虽然标准的FL配方捕捉无约束或全局约束的问题，许多实际的设置涉及异构资源或模型的约束，导致优化问题与代理特定的可行集。在这里，我们研究了一个个性化的约束联邦优化问题，其中每个代理与一个凸的局部目标和一个私人约束集。我们提出了PC-FedAvg，一种方法，在该方法中，每个代理保持交叉估计的其他代理的变量，通过多块本地决策向量。每个代理在本地更新所有块，只在自己的块中惩罚不可行性。此外，交叉估计机制使个性化，而不需要共识或共享代理之间的约束信息。我们建立的通信复杂度率为$\mathcal{O}（ε^{-2}）$次优和$\mathcal{O}（ε^{-1}）$代理明智的不可行性。MNIST和CIFAR-10数据集上的初步实验验证了我们的理论研究结果。
摘要：Federated learning (FL) has emerged as a communication-efficient algorithmic framework for distributed learning across multiple agents. While standard FL formulations capture unconstrained or globally constrained problems, many practical settings involve heterogeneous resource or model constraints, leading to optimization problems with agent-specific feasible sets. Here, we study a personalized constrained federated optimization problem in which each agent is associated with a convex local objective and a private constraint set. We propose PC-FedAvg, a method in which each agent maintains cross-estimates of the other agents' variables through a multi-block local decision vector. Each agent updates all blocks locally, penalizing infeasibility only in its own block. Moreover, the cross-estimate mechanism enables personalization without requiring consensus or sharing constraint information among agents. We establish communication-complexity rates of $\mathcal{O}(ε^{-2})$ for suboptimality and $\mathcal{O}(ε^{-1})$ for agent-wise infeasibility. Preliminary experiments on the MNIST and CIFAR-10 datasets validate our theoretical findings.
【3】Scalable Cross-Facility Federated Learning for Scientific Foundation Models on Multiple Supercomputers
标题：多台超级计算机上科学基础模型的可扩展跨设施联邦学习
链接：https://arxiv.org/abs/2603.19544
摘要：用于科学应用的人工智能越来越需要在由于隐私限制、数据主权或生成的大量数据而无法集中的数据上训练大型模型。联邦学习（FL）通过在不集中原始数据的情况下实现协作训练来解决这个问题，但是科学应用程序需要模型扩展，这需要大量的计算资源，通常在高性能计算（HPC）设施中提供。在HPC设施中部署FL实验会带来超出云或企业设置的挑战。我们为异构HPC环境提供了一个全面的跨设施FL框架，该框架基于具有Globus计算和传输编排的高级隐私保护联邦学习（APPFL）框架，并在四台美国能源部（DOE）领导级超级计算机上进行评估。我们证明，FL实验在HPC设施实际上是可以实现的，表征的异质性影响训练性能的关键来源，并表明，算法的选择显着现实的HPC调度条件下。我们通过对化学指令数据集上的大型语言模型进行微调来验证科学适用性，并将智能算法设计确定为未来部署的关键开放挑战。
摘要 ：Artificial Intelligence for scientific applications increasingly requires training large models on data that cannot be centralized due to privacy constraints, data sovereignty, or the sheer volume of data generated. Federated learning (FL) addresses this by enabling collaborative training without centralizing raw data, but scientific applications demand model scales that requires extensive computing resources, typically offered at High Performance Computing (HPC) facilities. Deploying FL experiments across HPC facilities introduces challenges beyond cloud or enterprise settings. We present a comprehensive cross-facility FL framework for heterogeneous HPC environments, built on Advanced Privacy-Preserving Federated Learning (APPFL) framework with Globus Compute and Transfer orchestration, and evaluate it across four U.S. Department of Energy (DOE) leadership-class supercomputers. We demonstrate that FL experiments across HPC facilities are practically achievable, characterize key sources of heterogeneity impacting the training performance, and show that algorithmic choices matter significantly under realistic HPC scheduling conditions. We validate the scientific applicability by fine-tuning a large language model on a chemistry instruction dataset, and identify scheduler-aware algorithm design as a critical open challenge for future deployments.
推理|分析|理解|解释(10篇)
【1】FIPO: Eliciting Deep Reasoning with Future-KL Influenced Policy Optimization
标题：FIPO：用未来激发深度推理-KL影响的政策优化
链接：https://arxiv.org/abs/2603.19835
摘要：我们提出了Future-KL Influenced Policy Optimization（FIPO），这是一种强化学习算法，旨在克服大型语言模型中的推理瓶颈。虽然GRPO风格的训练可以有效地扩展，但它通常依赖于基于结果的奖励（ORM），该奖励在轨迹中的每个令牌上均匀分布全局优势。我们认为，这种粗粒度的信用分配强加了性能上限，未能区分关键的逻辑支点从平凡的令牌。FIPO通过将贴现的未来KL分歧纳入政策更新来解决这个问题，创建一个密集的优势公式，根据令牌对后续轨迹行为的影响重新加权令牌。从经验上讲，FIPO使模型能够突破标准基线中的长度停滞。在Qwen2.5- 32 B上进行评估，FIPO将平均思维链长度从大约4，000扩展到超过10，000个令牌，并将AIME 2024 Pass@1的准确率从50.0%提高到58.0%的峰值（收敛于约56.0%）。这优于DeepSeek-R1-Zero-Math-32 B（约47.0%）和o 1-mini（约56.0%）。我们的研究结果表明，建立密集的优势配方是一个重要的路径不断发展的ORM为基础的算法，以解锁基础模型的全部推理潜力。我们开源了我们的培训系统，建立在verl框架上。
摘要：We present Future-KL Influenced Policy Optimization (FIPO), a reinforcement learning algorithm designed to overcome reasoning bottlenecks in large language models. While GRPO style training scales effectively, it typically relies on outcome-based rewards (ORM) that distribute a global advantage uniformly across every token in a trajectory. We argue that this coarse-grained credit assignment imposes a performance ceiling by failing to distinguish critical logical pivots from trivial tokens. FIPO addresses this by incorporating discounted future-KL divergence into the policy update, creating a dense advantage formulation that re-weights tokens based on their influence on subsequent trajectory behavior. Empirically, FIPO enables models to break through the length stagnation seen in standard baselines. Evaluated on Qwen2.5-32B, FIPO extends the average chain-of-thought length from roughly 4,000 to over 10,000 tokens and increases AIME 2024 Pass@1 accuracy from 50.0% to a peak of 58.0% (converging at approximately 56.0\%). This outperforms both DeepSeek-R1-Zero-Math-32B (around 47.0%) and o1-mini (approximately 56.0%). Our results suggest that establishing dense advantage formulations is a vital path for evolving ORM-based algorithms to unlock the full reasoning potential of base models. We open-source our training system, built on the verl framework.
【2】Regret Analysis of Sleeping Competing Bandits
标题：沉睡竞争盗贼的遗憾分析
链接：https://arxiv.org/abs/2603.19700
备注：29 pages, 3 figures
摘要：竞争强盗框架是最近新兴的领域，它将在线学习中的多武装强盗与博弈论中的稳定匹配相结合。虽然传统模型假设所有参与者和武器都是持续可用的，但在现实世界的问题中，他们的可用性可以随时间任意变化。在本文中，我们将这种设置表述为睡觉的竞争强盗。为了分析这个问题，我们自然地扩展了现有的竞争土匪的后悔定义，并得出建议的模型的后悔界。我们提出了一个算法，在合理的假设下，同时达到了$\mathrm{O}\left（NK\log T_{i}/Δ^2\right）$的渐近后悔界，其中$N$是参与者的数量，$K$是手臂的数量，$T_{i}$是每个参与者的回合数$p_i$，$Δ$是最小奖励差距。在同样的假设下，我们还给出了一个遗憾下界$\mathrmΩ\left（N（K-N+1）\log T_{i}/Δ^2 \right）$.这意味着我们的算法是渐近最优的政权的武器$K$的数量相对大于玩家的数量$N$。
摘要：The Competing Bandits framework is a recently emerging area that integrates multi-armed bandits in online learning with stable matching in game theory. While conventional models assume that all players and arms are constantly available, in real-world problems, their availability can vary arbitrarily over time. In this paper, we formulate this setting as Sleeping Competing Bandits. To analyze this problem, we naturally extend the regret definition used in existing competing bandits and derive regret bounds for the proposed model. We propose an algorithm that simultaneously achieves an asymptotic regret bound of $\mathrm{O}\left(NK\log T_{i}/Δ^2\right)$ under reasonable assumptions, where $N$ is the number of players, $K$ is the number of arms, $T_{i}$ is the number of rounds of each player $p_i$, and $Δ$ is the minimum reward gap. We also provide a regret lower bound of $\mathrmΩ\left( N(K-N+1)\log T_{i}/Δ^2 \right)$ under the same assumptions. This implies that our algorithm is asymptotically optimal in the regime where the number of arms $K$ is relatively larger than the number of players $N$.
【3】Ensembles-based Feature Guided Analysis
标题：基于集成的特征引导分析
链接：https://arxiv.org/abs/2603.19653
摘要：最近的深度神经网络（DNN）应用需要能够解释其行为的技术。诸如特征引导分析（FGA）的现有解决方案提取关于其内部行为的规则，例如，通过提供与神经元激活相关的解释。来自文献的结果表明，这些规则具有相当高的精度（即，它们正确地预测某些类别的特征），但是召回（即，这些规则适用的情况数量）更加有限。为了解决这个问题，本文提出了基于语义的特征引导分析（EFGA）。EFGA将FGA提取的规则组合成集成。集成根据聚合标准聚合不同的规则以增加其适用性，聚合标准是一种规定如何将规则组合成集成的策略。尽管我们的解决方案是可扩展的，并且用户可以开发不同的聚合标准，但在这项工作中，我们考虑了三种不同的聚合标准。我们评估了标准的选择如何影响EFGA在两个基准上的有效性（即，MNIST和LSC数据集），并发现不同的聚合标准提供了精确度和召回率之间的替代权衡。然后我们比较EFGA和FGA。对于这个实验，我们选择了一个聚合标准，提供了一个合理的权衡精度和召回。实验结果表明，EFGA比FGA具有更高的训练召回率（MNIST为+28.51%，LSC为+33.15%）和测试召回率（MNIST为+25.76%，LSC为+30.81%），而测试准确率（MNIST为-0.89%，LSC为-0.69%）几乎没有降低。
摘要：Recent Deep Neural Networks (DNN) applications ask for techniques that can explain their behavior. Existing solutions, such as Feature Guided Analysis (FGA), extract rules on their internal behaviors, e.g., by providing explanations related to neurons activation. Results from the literature show that these rules have considerable precision (i.e., they correctly predict certain classes of features), but the recall (i.e., the number of situations these rule apply) is more limited. To mitigate this problem, this paper presents Ensembles-based Feature Guided Analysis (EFGA). EFGA combines rules extracted by FGA into ensembles. Ensembles aggregate different rules to increase their applicability depending on an aggregation criterion, a policy that dictates how to combine rules into ensembles. Although our solution is extensible, and different aggregation criteria can be developed by users, in this work, we considered three different aggregation criteria. We evaluated how the choice of the criterion influences the effectiveness of EFGA on two benchmarks (i.e., the MNIST and LSC datasets), and found that different aggregation criteria offer alternative trade-offs between precision and recall. We then compare EFGA with FGA. For this experiment, we selected an aggregation criterion that provides a reasonable trade-off between precision and recall. Our results show that EFGA has higher train recall (+28.51% on MNIST, +33.15% on LSC), and test recall (+25.76% on MNIST, +30.81% on LSC) than FGA, with a negligible reduction on the test precision (-0.89% on MNIST, -0.69% on LSC).
【4】Heavy-Tailed and Long-Range Dependent Noise in Stochastic Approximation: A Finite-Time Analysis
标题：随机逼近中的重尾和长期相关噪音：临时分析
链接：https://arxiv.org/abs/2603.19648
备注：Submitted to IEEE Transactions on Automatic Control
摘要：随机逼近（SA）是一个基本的迭代框架，在强化学习和优化中有着广泛的应用。经典的分析通常依赖于鞅差或马尔可夫噪声与有界的二阶矩，但许多实际的设置，包括金融和通信，经常遇到重尾和长程相关（LRD）噪声。在这项工作中，我们研究了SA在这些非经典噪声模型下寻找强单调算子的根。我们建立了第一个有限时间矩界在这两种设置，提供明确的收敛速度，量化的影响重尾和时间依赖。我们的分析采用了噪声平均参数，在不修改迭代的情况下正则化噪声的影响。最后，我们将我们的一般框架应用于随机梯度下降（SGD）和梯度播放，并通过数值实验证实我们的有限时间分析。
摘要：Stochastic approximation (SA) is a fundamental iterative framework with broad applications in reinforcement learning and optimization. Classical analyses typically rely on martingale difference or Markov noise with bounded second moments, but many practical settings, including finance and communications, frequently encounter heavy-tailed and long-range dependent (LRD) noise. In this work, we study SA for finding the root of a strongly monotone operator under these non-classical noise models. We establish the first finite-time moment bounds in both settings, providing explicit convergence rates that quantify the impact of heavy tails and temporal dependence. Our analysis employs a noise-averaging argument that regularizes the impact of noise without modifying the iteration. Finally, we apply our general framework to stochastic gradient descent (SGD) and gradient play, and corroborate our finite-time analysis through numerical experiments.
【5】Demonstrations, CoT, and Prompting: A Theoretical Analysis of ICL
标题：演示、CoT和预算：ICL的理论分析
链接：https://arxiv.org/abs/2603.19611
摘要：情境学习（ICL）使预训练的LLM能够通过以一小组输入输出演示为条件来适应下游任务，而无需任何参数更新。虽然已经有许多理论上的努力来解释ICL是如何工作的，但大多数要么依赖于强大的架构或数据假设，要么未能捕捉到关键实践因素的影响，如演示选择，思想链（CoT）提示，演示的数量和提示模板。我们通过建立一个温和的假设下，这些设计选择的泛化行为的ICL的理论分析，解决这个差距。我们推导出了ICL测试损失的上限，表明性能受以下因素的影响：（i）所选演示的质量，由连接测试提示和预训练样本的路径上ICL损失的Lipschitz常数量化，（ii）预训练模型的内在ICL能力，以及（iii）分布偏移的程度。在同一个框架内，我们分析了CoT提示诱导任务分解，并表明，这是有益的，当示范是很好地选择在每个子步骤和由此产生的子任务更容易学习。最后，我们的特点是如何ICL性能的敏感性，提示模板随示威的数量而变化。总之，我们的研究表明，预训练使模型具有超越观察到的任务的泛化能力，而CoT使模型能够将更简单的子任务组合成更复杂的子任务，演示和指令使其能够检索类似或复杂的任务，包括那些可以组合成更复杂的任务的任务，共同支持对看不见的任务的泛化。所有的理论见解都是由实验证实的。
摘要：In-Context Learning (ICL) enables pretrained LLMs to adapt to downstream tasks by conditioning on a small set of input-output demonstrations, without any parameter updates. Although there have been many theoretical efforts to explain how ICL works, most either rely on strong architectural or data assumptions, or fail to capture the impact of key practical factors such as demonstration selection, Chain-of-Thought (CoT) prompting, the number of demonstrations, and prompt templates. We address this gap by establishing a theoretical analysis of ICL under mild assumptions that links these design choices to generalization behavior. We derive an upper bound on the ICL test loss, showing that performance is governed by (i) the quality of selected demonstrations, quantified by Lipschitz constants of the ICL loss along paths connecting test prompts to pretraining samples, (ii) an intrinsic ICL capability of the pretrained model, and (iii) the degree of distribution shift. Within the same framework, we analyze CoT prompting as inducing a task decomposition and show that it is beneficial when demonstrations are well chosen at each substep and the resulting subtasks are easier to learn. Finally, we characterize how ICL performance sensitivity to prompt templates varies with the number of demonstrations. Together, our study shows that pretraining equips the model with the ability to generalize beyond observed tasks, while CoT enables the model to compose simpler subtasks into more complex ones, and demonstrations and instructions enable it to retrieve similar or complex tasks, including those that can be composed into more complex ones, jointly supporting generalization to unseen tasks. All theoretical insights are corroborated by experiments.
【6】A Mathematical Theory of Understanding
标题：数学理解理论
链接：https://arxiv.org/abs/2603.19349
摘要：生成式人工智能已经改变了信息生产的经济学，以非常低的成本提供解释，证明，示例和分析。然而，信息的价值仍然取决于下游用户是否能够吸收并采取行动，信号只向具有解码能力的学习者传达意义：对一个用户来说，澄清概念的解释对于缺乏相关先决条件的另一个用户来说可能是噪音。本文提出了一个数学模型，学习者侧的瓶颈。我们将学习者建模为一个思维，一个抽象的学习系统，其特征在于概念上的先决结构。心智可以代表人类学习者、人工学习者（如神经网络）或任何解释信号的能力取决于先前获得的概念的代理。教学被建模为与潜在目标的顺序通信。因为只有当学习者已经获得了解析它们所需的先决条件时，教学信号才是可用的，所以有效的沟通渠道取决于学习者当前的知识状态，并且随着学习的进展而变得更加信息化。该模型产生了两个限制的学习和采用的速度：由先决条件的可达性和认识的目标的不确定性决定的结构限制。该框架暗示了培训和能力获取中的阈值效应。当教学视野低于目标的先决深度时，额外的教学不能成功地完成教学;一旦达到这个深度，完成就变得可行了。在不同的学习者中，普通的广播课程可能比个性化教学慢一个线性学习者类型数量的因素。
摘要：Generative AI has transformed the economics of information production, making explanations, proofs, examples, and analyses available at very low cost. Yet the value of information still depends on whether downstream users can absorb and act on it. A signal conveys meaning only to a learner with the structural capacity to decode it: an explanation that clarifies a concept for one user may be indistinguishable from noise to another who lacks the relevant prerequisites. This paper develops a mathematical model of that learner-side bottleneck. We model the learner as a mind, an abstract learning system characterized by a prerequisite structure over concepts. A mind may represent a human learner, an artificial learner such as a neural network, or any agent whose ability to interpret signals depends on previously acquired concepts. Teaching is modeled as sequential communication with a latent target. Because instructional signals are usable only when the learner has acquired the prerequisites needed to parse them, the effective communication channel depends on the learner's current state of knowledge and becomes more informative as learning progresses. The model yields two limits on the speed of learning and adoption: a structural limit determined by prerequisite reachability and an epistemic limit determined by uncertainty about the target. The framework implies threshold effects in training and capability acquisition. When the teaching horizon lies below the prerequisite depth of the target, additional instruction cannot produce successful completion of teaching; once that depth is reached, completion becomes feasible. Across heterogeneous learners, a common broadcast curriculum can be slower than personalized instruction by a factor linear in the number of learner types.
【7】FalconBC: Flow matching for Amortized inference of Latent-CONditioned physiologic Boundary Conditions
标题：FalconBC：用于潜伏条件生理边界条件的摊销推断的流匹配
链接：https://arxiv.org/abs/2603.19331
摘要：边界条件调整是特定于患者的心血管建模中的基本步骤。尽管离线训练成本增加，但最近的数据驱动变分推理方法可以有效地估计边界条件的联合后验分布，并在临床目标上摊销训练工作。然而，即使是最现代的方法在两个重要的场景中也有不足之处：具有已知平均流量和假设波形形状的开环模型，以及受血管病变影响的解剖结构，其中分割影响压力或流量分割目标的可达性。在这两种情况下，边界条件都不能单独调整。我们介绍了一个一般的摊销推理框架的基础上，概率流，治疗临床目标，流入功能，和点云嵌入的患者特定的解剖结构，无论是条件变量或数量联合估计。我们在两个患者特定模型上演示了该方法：狭窄位置和严重程度不同的髂动脉分叉和冠状动脉树。
摘要：Boundary condition tuning is a fundamental step in patient-specific cardiovascular modeling. Despite an increase in offline training cost, recent methods in data-driven variational inference can efficiently estimate the joint posterior distribution of boundary conditions, with amortization of training efforts over clinical targets. However, even the most modern approaches fall short in two important scenarios: open-loop models with known mean flow and assumed waveform shapes, and anatomies affected by vascular lesions where segmentation influences the reachability of pressure or flow split targets. In both cases, boundary conditions cannot be tuned in isolation. We introduce a general amortized inference framework based on probabilistic flow that treats clinical targets, inflow features, and point cloud embeddings of patient-specific anatomies as either conditioning variables or quantities to be jointly estimated. We demonstrate the approach on two patient-specific models: an aorto-iliac bifurcation with varying stenosis locations and severity, and a coronary arterial tree.
【8】VERDICT: Verifiable Evolving Reasoning with Directive-Informed Collegial Teams for Legal Judgment Prediction
标题：VERDICT：具有指导性知情学院团队的可验证的进化推理，用于法律判断预测
链接：https://arxiv.org/abs/2603.19306
备注：15 pages,3 figures,4 tables
摘要：法律判决预测是从案件事实中预测适用的法律条款、罪名和刑罚条款。除了准确性之外，LJP还要求本质上可解释的和有法律依据的推理，这些推理可以使法定规则与先例标准相协调。然而，现有的方法往往表现为静态的，一次性的预测，提供有限的程序支持可验证的推理和小的能力，以适应审慎的做法演变。我们提出了VERDICT，一个自我完善的协作多代理框架，模拟一个虚拟的合议庭。VERDICT将专门的代理人分配到互补的角色（例如，事实构建、法律检索、意见起草和监督验证），并在可追溯的草案-验证-修订工作流程中协调它们，并提供明确的通过/修改反馈，产生可验证的推理痕迹和修订依据。为了捕捉不断变化的案例经验，我们进一步引入了一个基于微指令范式的混合法理记忆（HJM），它存储了先例标准，并不断将经过验证的多代理验证轨迹提取到更新的微指令中，以便在案例中不断学习。我们在CAIL 2018和新构建的CJO 2025数据集上评估了VERDICT，该数据集具有严格的未来时间分割以进行时间概括。VERDICT在CAIL 2018上实现了最先进的性能，并在CJO 2025上展示了强大的泛化能力。为了促进可重复性和进一步的研究，我们在https://anonymous.4open.science/r/ARR-4437上发布了我们的代码和数据集。
摘要：Legal Judgment Prediction (LJP) predicts applicable law articles, charges, and penalty terms from case facts. Beyond accuracy, LJP calls for intrinsically interpretable and legally grounded reasoning that can reconcile statutory rules with precedent-informed standards. However, existing methods often behave as static, one-shot predictors, providing limited procedural support for verifiable reasoning and little capability to adapt as jurisprudential practice evolves. We propose VERDICT, a self-refining collaborative multi-agent framework that simulates a virtual collegial panel. VERDICT assigns specialized agents to complementary roles (e.g., fact structuring, legal retrieval, opinion drafting, and supervisory verification) and coordinates them in a traceable draft--verify--revise workflow with explicit Pass/Reject feedback, producing verifiable reasoning traces and revision rationales. To capture evolving case experience, we further introduce a Hybrid Jurisprudential Memory (HJM) grounded in the Micro-Directive Paradigm, which stores precedent standards and continually distills validated multi-agent verification trajectories into updated Micro-Directives for continual learning across cases. We evaluate VERDICT on CAIL2018 and a newly constructed CJO2025 dataset with a strict future time-split for temporal generalization. VERDICT achieves state-of-the-art performance on CAIL2018 and demonstrates strong generalization on CJO2025. To facilitate reproducibility and further research, we release our code and the dataset at https://anonymous.4open.science/r/ARR-4437.
【9】A Visualization for Comparative Analysis of Regression Models
标题：回归模型比较分析的可视化
链接：https://arxiv.org/abs/2603.19291
摘要：由于回归是一个被广泛研究的问题，已经提出了许多方法来解决它，每一种方法往往需要设置不同的超参数。因此，为给定的应用选择合适的方法可能非常困难，并且依赖于比较它们的性能。性能通常使用各种度量来衡量，例如平均绝对误差（MAE），均方根误差（RMSE）或R平方（R${}^2$）。这些指标通过量化预测值和实际值之间的差异，提供了预测准确性的数值总结。然而，尽管这些指标在文献中广泛用于总结模型性能，并有助于区分模型的表现好坏，但它们通常聚集了太多的信息。本文通过引入一种新颖的可视化方法来解决这些限制，该方法强调了回归模型性能的关键方面。所提出的方法建立在三个主要贡献之上：（1）考虑2D空间中的残差，这允许同时评估来自两个模型的误差，（2）利用Mahalanobis距离来解释数据内的相关性和尺度差异，以及（3）采用色图来可视化基于颜色的误差分布，从而更容易识别密集区域和离群值。通过图形化地表示错误的分布及其相关性，这种方法提供了一个更详细、更全面的模型性能视图，使用户能够发现传统聚合指标可能会掩盖的模式。所提出的可视化方法有助于更深入地了解回归模型的性能差异和误差分布，增强评估和比较过程。
摘要：As regression is a widely studied problem, many methods have been proposed to solve it, each of them often requiring setting different hyper-parameters. Therefore, selecting the proper method for a given application may be very difficult and relies on comparing their performances. Performance is usually measured using various metrics such as Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), or R-squared (R${}^2$). These metrics provide a numerical summary of predictive accuracy by quantifying the difference between predicted and actual values. However, while these metrics are widely used in the literature for summarizing model performance and useful to distinguish between models performing poorly and well, they often aggregate too much information. This article addresses these limitations by introducing a novel visualization approach that highlights key aspects of regression model performance. The proposed method builds upon three main contributions: (1) considering the residuals in a 2D space, which allows for simultaneous evaluation of errors from two models, (2) leveraging the Mahalanobis distance to account for correlations and differences in scale within the data, and (3) employing a colormap to visualize the percentile-based distribution of errors, making it easier to identify dense regions and outliers. By graphically representing the distribution of errors and their correlations, this approach provides a more detailed and comprehensive view of model performance, enabling users to uncover patterns that traditional aggregate metrics may obscure. The proposed visualization method facilitates a deeper understanding of regression model performance differences and error distributions, enhancing the evaluation and comparison process.
【10】Speculating Experts Accelerates Inference for Mixture-of-Experts
标题：猜测专家加速了对混合专家的推理
链接：https://arxiv.org/abs/2603.19289
摘要：混合专家（MoE）模型作为一种扩展大型语言模型（LLM）容量的方法，同时保持稀疏激活和减少每令牌计算，已经越来越受欢迎。然而，在内存受限的推理设置中，专家权重必须卸载到CPU，从而在解码期间产生来自CPU-GPU传输的性能瓶颈。我们提出了一个专家预取方案，利用目前计算的内部模型表示推测未来的专家，使内存传输与计算重叠。在多个MoE架构，我们证明，未来的专家可以可靠地预测这些内部表示。我们还表明，执行推测的专家一般保持下游任务的准确性，从而通过消除需要重新获取真正的路由器选择的专家保留更有效的计算内存重叠。与从CPU内存按需加载专家相比，我们的方法集成到优化的推理引擎中，每个输出令牌（TPOT）的时间减少了高达14%。对于MoEs，投机执行单独产生次优的准确性，我们进一步研究轻量级的估计，提高专家预测命中率，从而减少性能下降。我们的代码在https://github.com/axonn-ai/yalis/tree/offload_prefetch上开源发布。
摘要：Mixture-of-Experts (MoE) models have gained popularity as a means of scaling the capacity of large language models (LLMs) while maintaining sparse activations and reduced per-token compute. However, in memory-constrained inference settings, expert weights must be offloaded to CPU, creating a performance bottleneck from CPU-GPU transfers during decoding. We propose an expert prefetching scheme that leverages currently computed internal model representations to speculate future experts, enabling memory transfers to overlap with computation. Across multiple MoE architectures, we demonstrate that future experts can be reliably predicted by these internal representations. We also demonstrate that executing speculated experts generally maintains downstream task accuracy, thus preserving more effective compute-memory overlap by eliminating the need to re-fetch true router-selected experts. Integrated into an optimized inference engine, our approach achieves up to 14\% reduction in time per output token (TPOT) over on-demand loading of experts from CPU memory. For MoEs where speculative execution alone yields suboptimal accuracy, we further examine lightweight estimators that improve expert prediction hit rates, thereby reducing performance degradation. Our code is released in open-source at https://github.com/axonn-ai/yalis/tree/offload_prefetch.
检测相关(1篇)
【1】ODySSeI: An Open-Source End-to-End Framework for Automated Detection, Segmentation, and Severity Estimation of Lesions in Invasive Coronary Angiography Images
标题：OdySSeI：用于自动检测、分割和严重性估计有创冠状动脉血管造影图像中病变的开源端到端框架
链接：https://arxiv.org/abs/2603.20021
摘要：有创性冠状动脉造影（ICA）是评估冠状动脉疾病的临床金标准。然而，它的解释仍然是主观的，并倾向于内部和操作员之间的变化。在这项工作中，我们介绍了ODySSeI：一个开源的端到端框架，用于ICA图像中病变的自动检测，分割和严重程度估计。ODySSeI集成了基于深度学习的病变检测和使用新型金字塔增强方案（PAS）训练的病变分割模型，以增强不同患者队列（来自欧洲，北美和亚洲的2149名患者）的鲁棒性和实时性能。此外，我们提出了一种定量冠状动脉造影病变严重程度估计（LSE）技术，直接计算最小管腔直径（MLD）和直径狭窄的预测病变的几何形状。对分布内和分布外临床数据集的广泛评价证明了ODySSeI的强大通用性。与相对简单的任务相比，我们的PAS在高度复杂的任务中产生了很大的性能增益，值得注意的是，病变检测性能提高了2.5倍，而病变分割性能提高了1- 3%。我们的LSE技术实现了高精度，预测的MLD值仅与相应的地面实况相差$\pm $2 -3个像素。平均而言，ODySSeI在CPU上仅需几秒钟即可处理原始ICA图像，在GPU上只需几分之一秒，并可在swisscardia.epfl. ch上作为即插即用的Web界面提供。总体而言，这项工作将ODySSeI确立为一个全面的开源框架，支持自动化，可重复和可扩展的ICA分析，用于实时临床决策。
摘要：Invasive Coronary Angiography (ICA) is the clinical gold standard for the assessment of coronary artery disease. However, its interpretation remains subjective and prone to intra- and inter-operator variability. In this work, we introduce ODySSeI: an Open-source end-to-end framework for automated Detection, Segmentation, and Severity estimation of lesions in ICA images. ODySSeI integrates deep learning-based lesion detection and lesion segmentation models trained using a novel Pyramidal Augmentation Scheme (PAS) to enhance robustness and real-time performance across diverse patient cohorts (2149 patients from Europe, North America, and Asia). Furthermore, we propose a quantitative coronary angiography-free Lesion Severity Estimation (LSE) technique that directly computes the Minimum Lumen Diameter (MLD) and diameter stenosis from the predicted lesion geometry. Extensive evaluation on both in-distribution and out-of-distribution clinical datasets demonstrates ODySSeI's strong generalizability. Our PAS yields large performance gains in highly complex tasks as compared to relatively simpler ones, notably, a 2.5-fold increase in lesion detection performance versus a 1-3\% increase in lesion segmentation performance over their respective baselines. Our LSE technique achieves high accuracy, with predicted MLD values differing by only $\pm$ 2-3 pixels from the corresponding ground truths. On average, ODySSeI processes a raw ICA image within only a few seconds on a CPU and in a fraction of a second on a GPU and is available as a plug-and-play web interface at swisscardia.epfl.ch. Overall, this work establishes ODySSeI as a comprehensive and open-source framework which supports automated, reproducible, and scalable ICA analysis for real-time clinical decision-making.
分类|识别(4篇)
【1】Enhancing Hyperspace Analogue to Language (HAL) Representations via Attention-Based Pooling for Text Classification
标题：通过基于注意力的池化文本分类增强超空间对语言（HAL）表示的类比
链接：https://arxiv.org/abs/2603.20149
备注：7 pages, 1 figure, 1 table
摘要：超空间语义语言模型（HAL）依赖于全局词共现矩阵来构建分布式语义表示。虽然这些表示有效地捕获词汇关系，但通过标准均值池将它们聚合到嵌入级嵌入中通常会导致信息丢失。均值池为所有标记分配相同的权重，从而稀释了具有无信息结构标记的上下文突出词的影响。在本文中，我们通过将一个可学习的，温度缩放的添加剂注意力机制集成到HAL表示管道中来解决这个限制。为了减轻原始共生矩阵的稀疏性和高维性，我们应用截断奇异值分解（SVD）将向量投影到注意力层之前的密集潜在空间中。我们在IMDB情感分析数据集上评估了所提出的架构。实证结果表明，基于注意力池的方法达到了82.38%的测试准确率，产生了6.74个百分点的绝对提高比传统的均值池基线（75.64%）。此外，定性分析的注意力权重表明，该机制成功地抑制停用词，并有选择地参加情感承载令牌，提高分类性能和模型的可解释性。
摘要：The Hyperspace Analogue to Language (HAL) model relies on global word co-occurrence matrices to construct distributional semantic representations. While these representations capture lexical relationships effectively, aggregating them into sentence-level embeddings via standard mean pooling often results in information loss. Mean pooling assigns equal weight to all tokens, thereby diluting the impact of contextually salient words with uninformative structural tokens. In this paper, we address this limitation by integrating a learnable, temperature-scaled additive attention mechanism into the HAL representation pipeline. To mitigate the sparsity and high dimensionality of the raw co-occurrence matrices, we apply Truncated Singular Value Decomposition (SVD) to project the vectors into a dense latent space prior to the attention layer. We evaluate the proposed architecture on the IMDB sentiment analysis dataset. Empirical results demonstrate that the attention-based pooling approach achieves a test accuracy of 82.38%, yielding an absolute improvement of 6.74 percentage points over the traditional mean pooling baseline (75.64%). Furthermore, qualitative analysis of the attention weights indicates that the mechanism successfully suppresses stop-words and selectively attends to sentiment-bearing tokens, improving both classification performance and model interpretability.
【2】Continual Learning for Food Category Classification Dataset: Enhancing Model Adaptability and Performance
标题：食品类别分类数据集的持续学习：增强模型适应性和性能
链接：https://arxiv.org/abs/2603.19624
摘要：传统的机器学习管道通常很难识别原始训练集中没有的类别。这种差距通常会降低准确性，因为固定的数据集很少能捕捉到域的全部多样性。为了解决这个问题，我们提出了一个持续学习的框架文本引导的食品分类。与需要从头开始重新训练的方法不同，我们的方法支持增量更新，允许在不降低先验知识的情况下整合新类别。例如，一个在西餐上训练过的模型可以在以后学习对dosa或泡菜等菜肴进行分类。虽然还需要进一步的改进，但这种设计显示出适应性食物识别的前景，并可应用于饮食监测和个性化营养计划。
摘要 ：Conventional machine learning pipelines often struggle to recognize categories absent from the original trainingset. This gap typically reduces accuracy, as fixed datasets rarely capture the full diversity of a domain. To address this, we propose a continual learning framework for text-guided food classification. Unlike approaches that require retraining from scratch, our method enables incremental updates, allowing new categories to be integrated without degrading prior knowledge. For example, a model trained on Western cuisines could later learn to classify dishes such as dosa or kimchi. Although further refinements are needed, this design shows promise for adaptive food recognition, with applications in dietary monitoring and personalized nutrition planning.
【3】PFM-VEPAR: Prompting Foundation Models for RGB-Event Camera based Pedestrian Attribute Recognition
标题：PFM-VEBAR：基于RGB事件摄像机的行人属性识别的预确定基础模型
链接：https://arxiv.org/abs/2603.19565
摘要：基于事件的行人属性识别（PAR）利用运动线索来增强低光和运动模糊场景中的RGB摄像头，从而能够更准确地推断年龄和情感等属性。然而，现有的双流多模态融合方法引入了显着的计算开销，忽略了有价值的指导，从上下文样本。为了解决这些局限性，本文提出了一个事件过滤器。抛弃了计算上昂贵的辅助主干，该模块直接将极其轻量级和高效的离散余弦变换（DCT）和逆DCT（IDCT）操作应用于事件数据。该设计以最小的计算成本提取频域事件特征，从而有效地增强RGB分支。此外，设计用于提供丰富先验知识的外部存储库与现代Hopfield网络相结合，可以实现联想记忆增强表示学习。这种机制有效地挖掘和利用了不同样本之间的全局关系知识。最后，交叉注意机制融合了RGB和事件模态，然后通过前馈网络进行属性预测。在多个基准数据集上的大量实验充分验证了所提出的RGB-Event PAR框架的有效性。本文的源代码将在https://github.com/Event-AHU/OpenPAR上发布
摘要：Event-based pedestrian attribute recognition (PAR) leverages motion cues to enhance RGB cameras in low-light and motion-blur scenarios, enabling more accurate inference of attributes like age and emotion. However, existing two-stream multimodal fusion methods introduce significant computational overhead and neglect the valuable guidance from contextual samples. To address these limitations, this paper proposes an Event Prompter. Discarding the computationally expensive auxiliary backbone, this module directly applies extremely lightweight and efficient Discrete Cosine Transform (DCT) and Inverse DCT (IDCT) operations to the event data. This design extracts frequency-domain event features at a minimal computational cost, thereby effectively augmenting the RGB branch. Furthermore, an external memory bank designed to provide rich prior knowledge, combined with modern Hopfield networks, enables associative memory-augmented representation learning. This mechanism effectively mines and leverages global relational knowledge across different samples. Finally, a cross-attention mechanism fuses the RGB and event modalities, followed by feed-forward networks for attribute prediction. Extensive experiments on multiple benchmark datasets fully validate the effectiveness of the proposed RGB-Event PAR framework. The source code of this paper will be released on https://github.com/Event-AHU/OpenPAR
【4】MSNet and LS-Net: Scalable Multi-Scale Multi-Representation Networks for Time Series Classification
标题：MSNet和LS-Net：用于时间序列分类的可扩展多尺度多表示网络
链接：https://arxiv.org/abs/2603.19315
摘要：时间序列分类（TSC）的性能不仅取决于体系结构设计，还取决于输入表示的多样性。在这项工作中，我们提出了一个可扩展的多尺度卷积框架，该框架系统地集成了单变量时间序列的结构化多表示输入。 我们介绍两种架构：MSNet是一种针对鲁棒性和校准进行优化的分层多尺度卷积网络，LS-Net是一种为效率感知部署而设计的轻量级变体。此外，我们调整LiteMV -最初为多变量输入开发-对多表示单变量信号进行操作，从而实现交叉表示交互。 我们在统一的实验协议下评估了142个基准数据集的所有模型。关键差异分析证实了顶级模型之间的统计学显著性能差异。结果表明，LiteMV实现了最高的平均精度，MSNet提供了卓越的概率校准（最低的NLL），LS-Net提供了最佳的效率-精度权衡。帕累托分析进一步表明，多表示多尺度建模产生了灵活的设计空间，可以针对面向精度、面向校准或资源受限的设置进行调整。 这些研究结果建立了可扩展的多表征多尺度学习作为现代TSC的原则和实用方向。MSNet和LS-Net的参考实现可在https://github.com/alagoz/msnet-lsnet-tsc上获得
摘要：Time series classification (TSC) performance depends not only on architectural design but also on the diversity of input representations. In this work, we propose a scalable multi-scale convolutional framework that systematically integrates structured multi-representation inputs for univariate time series. We introduce two architectures: MSNet, a hierarchical multi-scale convolutional network optimized for robustness and calibration, and LS-Net, a lightweight variant designed for efficiency-aware deployment. In addition, we adapt LiteMV -- originally developed for multivariate inputs -- to operate on multi-representation univariate signals, enabling cross-representation interaction. We evaluate all models across 142 benchmark datasets under a unified experimental protocol. Critical Difference analysis confirms statistically significant performance differences among the top models. Results show that LiteMV achieves the highest mean accuracy, MSNet provides superior probabilistic calibration (lowest NLL), and LS-Net offers the best efficiency-accuracy tradeoff. Pareto analysis further demonstrates that multi-representation multi-scale modeling yields a flexible design space that can be tuned for accuracy-oriented, calibration-oriented, or resource-constrained settings. These findings establish scalable multi-representation multi-scale learning as a principled and practical direction for modern TSC. Reference implementation of MSNet and LS-Net is available at: https://github.com/alagoz/msnet-lsnet-tsc
表征(2篇)
【1】Spectral Alignment in Forward-Backward Representations via Temporal Abstraction
标题：通过时间抽象实现前向-后向表示中的谱对齐
链接：https://arxiv.org/abs/2603.20103
摘要：前向后（FB）表示提供了一个强大的框架，学习连续空间中的后继表示（SR）通过强制执行低秩分解。然而，一个基本的频谱失配之间经常存在的连续环境的高秩过渡动力学和低秩的FB架构的瓶颈，使准确的低秩表示学习困难。在这项工作中，我们分析时间抽象的机制，以减轻这种不匹配。通过表征的过渡算子的频谱特性，我们表明，时间抽象作为一个低通滤波器，抑制高频频谱分量。这种抑制降低了诱导SR的有效秩，同时保持对所得值函数误差的正式界。从经验上讲，我们表明，这种对齐是稳定FB学习的关键因素，特别是在高折扣因素，引导变得容易出错。我们的研究结果确定时间抽象作为一个原则性的机制，用于塑造底层MDP的频谱结构，并使有效的长期表示在连续控制。
摘要：Forward-backward (FB) representations provide a powerful framework for learning the successor representation (SR) in continuous spaces by enforcing a low-rank factorization. However, a fundamental spectral mismatch often exists between the high-rank transition dynamics of continuous environments and the low-rank bottleneck of the FB architecture, making accurate low-rank representation learning difficult. In this work, we analyze temporal abstraction as a mechanism to mitigate this mismatch. By characterizing the spectral properties of the transition operator, we show that temporal abstraction acts as a low-pass filter that suppresses high-frequency spectral components. This suppression reduces the effective rank of the induced SR while preserving a formal bound on the resulting value function error. Empirically, we show that this alignment is a key factor for stable FB learning, particularly at high discount factors where bootstrapping becomes error-prone. Our results identify temporal abstraction as a principled mechanism for shaping the spectral structure of the underlying MDP and enabling effective long-horizon representations in continuous control.
【2】RiboSphere: Learning Unified and Efficient Representations of RNA Structures
标题：RiboSphere：学习RNA结构的统一和有效表示
链接
：https://arxiv.org/abs/2603.19636
摘要：精确的RNA结构建模仍然很困难，因为RNA骨架是高度灵活的，非典型的相互作用是普遍的，实验确定的3D结构是相对稀缺的。我们介绍\n {RiboSphere}，一个框架，学习\n {离散}几何表示的RNA相结合的矢量量化与流匹配。我们的设计受到RNA结构的模块化组织的启发：复杂的折叠由重复的结构基序组成。RiboSphere使用几何Transformer编码器来产生SE（3）不变（旋转/平移不变）特征，这些特征通过有限标量量化（FSQ）离散化为有限的潜在代码词汇表。以这些离散代码为条件，流匹配解码器重建原子坐标，从而实现高保真结构生成。我们发现，学习的代码索引富集特定的RNA基序，这表明该模型捕获基序水平的组成结构，而不是作为一个纯粹的压缩瓶颈。在基准测试中，RiboSphere在结构重建方面表现出色（RMSD 1.25\，RISK，TM评分0.84），其预训练的离散表示有效地转移到逆折叠和RNA-配体结合预测，在数据稀缺的情况下具有强大的泛化能力。
摘要：Accurate RNA structure modeling remains difficult because RNA backbones are highly flexible, non-canonical interactions are prevalent, and experimentally determined 3D structures are comparatively scarce. We introduce \emph{RiboSphere}, a framework that learns \emph{discrete} geometric representations of RNA by combining vector quantization with flow matching. Our design is motivated by the modular organization of RNA architecture: complex folds are composed from recurring structural motifs. RiboSphere uses a geometric transformer encoder to produce SE(3)-invariant (rotation/translation-invariant) features, which are discretized with finite scalar quantization (FSQ) into a finite vocabulary of latent codes. Conditioned on these discrete codes, a flow-matching decoder reconstructs atomic coordinates, enabling high-fidelity structure generation. We find that the learned code indices are enriched for specific RNA motifs, suggesting that the model captures motif-level compositional structure rather than acting as a purely compressive bottleneck. Across benchmarks, RiboSphere achieves strong performance in structure reconstruction (RMSD 1.25\,Å, TM-score 0.84), and its pretrained discrete representations transfer effectively to inverse folding and RNA--ligand binding prediction, with robust generalization in data-scarce regimes.
优化|敛散性(6篇)
【1】Quantifying Gate Contribution in Quantum Feature Maps for Scalable Circuit Optimization
标题：量化量子特征图中的门贡献以实现可扩展电路优化
链接：https://arxiv.org/abs/2603.19805
摘要：量子机器学习为分类任务提供了有前途的优势，但当前设备中的噪声、去相干和连接性约束继续限制基于特征图的电路的有效执行。门评估和阈值评估（GATE）是一种电路优化方法，减少量子特征图使用一种新的门的重要性指数。该指数通过结合保真度、纠缠度和灵敏度来量化每个门的相关性。它是制定为模拟器/仿真器环境，其中量子状态是可访问的，并为真正的硬件，其中这些数量估计从测量结果和辅助电路。该方法迭代地扫描阈值范围，消除低贡献门，生成优化的量子机器学习模型，并在最终测试之前根据准确性，运行时间和平衡的性能标准对其进行排名。该方法使用两个代表性的量子机器学习模型PegasosQSVM和量子神经网络在三个执行场景中对真实世界的分类数据集进行评估：无噪声仿真，来自IBM后端的噪声仿真和真实的IBM量子硬件。研究了特征图中门移除的结构影响，研究了与降噪技术的兼容性，并使用基于密度矩阵，矩阵乘积状态，张量网络和真实设备的方法评估了索引计算的可扩展性。结果表明，电路尺寸和运行时间的一致减少，在许多情况下，保留或提高预测精度，最佳的权衡通常发生在中间阈值，而不是在基线电路或那些压缩更积极。
摘要：Quantum machine learning offers promising advantages for classification tasks, but noise, decoherence, and connectivity constraints in current devices continue to limit the efficient execution of feature map-based circuits. Gate Assessment and Threshold Evaluation (GATE) is presented as a circuit optimization methodology that reduces quantum feature maps using a novel gate significance index. This index quantifies the relevance of each gate by combining fidelity, entanglement, and sensitivity. It is formulated for both simulator/emulator environments, where quantum states are accessible, and for real hardware, where these quantities are estimated from measurement results and auxiliary circuits. The approach iteratively scans a threshold range, eliminates low-contribution gates, generates optimized quantum machine learning models, and ranks them based on accuracy, runtime, and a balanced performance criterion before final testing. The methodology is evaluated on real-world classification datasets using two representative quantum machine learning models, PegasosQSVM and Quantum Neural Network, in three execution scenarios: noise-free simulation, noisy emulation derived from an IBM backend, and real IBM quantum hardware. The structural impact of gate removal in feature maps is examined, compatibility with noise-mitigation techniques is studied, and the scalability of index computation is evaluated using approaches based on density matrices, matrix product states, tensor networks, and real-world devices. The results show consistent reductions in circuit size and runtime and, in many cases, preserved or improved predictive accuracy, with the best trade-offs typically occurring at intermediate thresholds rather than in the baseline circuits or in those compressed more aggressively.
【2】Global Convergence of Multiplicative Updates for the Matrix Mechanism: A Collaborative Proof with Gemini 3
标题：矩阵机制乘性更新的全球收敛：Gemini 3的合作证明
链接：https://arxiv.org/abs/2603.19465
备注：12 pages, 1 figure
摘要：We analyze a fixed-point iteration $v \leftarrow φ(v)$ arising in the optimization of a regularized nuclear norm objective involving the Hadamard product structure, posed in~\cite{denisov} in the context of an optimization problem over the space of algorithms in private machine learning. We prove that the iteration $v^{(k+1)} = \text{diag}((D_{v^{(k)}}^{1/2} M D_{v^{(k)}}^{1/2})^{1/2})$ converges monotonically to the unique global optimizer of the potential function $J(v) = 2 \text{Tr}((D_v^{1/2} M D_v^{1/2})^{1/2}) - \sum v_i$, closing a problem left open there. The bulk of this proof was provided by Gemini 3, subject to some corrections and interventions. Gemini 3 also sketched the initial version of this note. Thus, it represents as much a commentary on the practical use of AI in mathematics as it represents the closure of a small gap in the literature. As such, we include a small narrative description of the prompting process, and some resulting principles for working with AI to prove mathematics.
摘要：We analyze a fixed-point iteration $v \leftarrow φ(v)$ arising in the optimization of a regularized nuclear norm objective involving the Hadamard product structure, posed in~\cite{denisov} in the context of an optimization problem over the space of algorithms in private machine learning. We prove that the iteration $v^{(k+1)} = \text{diag}((D_{v^{(k)}}^{1/2} M D_{v^{(k)}}^{1/2})^{1/2})$ converges monotonically to the unique global optimizer of the potential function $J(v) = 2 \text{Tr}((D_v^{1/2} M D_v^{1/2})^{1/2}) - \sum v_i$, closing a problem left open there. The bulk of this proof was provided by Gemini 3, subject to some corrections and interventions. Gemini 3 also sketched the initial version of this note. Thus, it represents as much a commentary on the practical use of AI in mathematics as it represents the closure of a small gap in the literature. As such, we include a small narrative description of the prompting process, and some resulting principles for working with AI to prove mathematics.
【3】Deep Hilbert--Galerkin Methods for Infinite-Dimensional PDEs and Optimal Control
标题：无限维偏置方程和最优控制的深Hilbert-Galerkin方法
链接：https://arxiv.org/abs/2603.19463
摘要：我们开发了基于深度学习的近似方法，用于可分离希尔伯特空间上的完全非线性二阶偏微分方程，例如用于无限维控制的HJB方程，通过希尔伯特-伽辽金神经算子（HGNOs）参数化解决方案。我们证明了第一个通用逼近定理（UAT），这是足够强大的，以解决这些问题，基于新的拓扑结构的Hessian条款和相应的新的连续性假设的完全非线性算子。这些拓扑是非顺序的和不可度量的，使得问题变得微妙。特别是，我们证明了希尔伯特空间上函数的UAT，连同它们的Fréchet导数到二阶，以及应用于一阶导数的无界算子，确保HGNO能够近似所有PDE项。对于控制问题，我们进一步证明了UAT的最优反馈控制在我们的近似值函数HGNO。 我们为这些问题开发了数值训练方法，我们称之为深度希尔伯特-伽辽金和希尔伯特Actor-Critic（强化学习）方法，通过最小化整个希尔伯特空间上PDE残差的$L^2_μ（H）$-范数，而不仅仅是投影到有限维的PDE。这是第一篇提出这种方法的论文。所考虑的模型出现在许多应用科学中，如物理学中的泛函微分方程和Kolmogorov和HJB偏微分方程，与受控偏微分方程、随机偏微分方程、路径依赖系统、部分观测随机系统和平均场偏微分方程有关。我们数值求解了与确定性和随机热方程和Burgers方程的最优控制相关的Kolmogorov和HJB偏微分方程的示例，展示了我们基于深度学习的方法的前景。
摘要：We develop deep learning-based approximation methods for fully nonlinear second-order PDEs on separable Hilbert spaces, such as HJB equations for infinite-dimensional control, by parameterizing solutions via Hilbert--Galerkin Neural Operators (HGNOs). We prove the first Universal Approximation Theorems (UATs) which are sufficiently powerful to address these problems, based on novel topologies for Hessian terms and corresponding novel continuity assumptions on the fully nonlinear operator. These topologies are non-sequential and non-metrizable, making the problem delicate. In particular, we prove UATs for functions on Hilbert spaces, together with their Fréchet derivatives up to second order, and for unbounded operators applied to the first derivative, ensuring that HGNOs are able to approximate all the PDE terms. For control problems, we further prove UATs for optimal feedback controls in terms of our approximating value function HGNO. We develop numerical training methods, which we call Deep Hilbert--Galerkin and Hilbert Actor-Critic (reinforcement learning) Methods, for these problems by minimizing the $L^2_μ(H)$-norm of the residual of the PDE on the whole Hilbert space, not just a projected PDE to finite dimensions. This is the first paper to propose such an approach. The models considered arise in many applied sciences, such as functional differential equations in physics and Kolmogorov and HJB PDEs related to controlled PDEs, SPDEs, path-dependent systems, partially observed stochastic systems, and mean-field SDEs. We numerically solve examples of Kolmogorov and HJB PDEs related to the optimal control of deterministic and stochastic heat and Burgers' equations, demonstrating the promise of our deep learning-based approach.
【4】Bridging Conformal Prediction and Scenario Optimization: Discarded Constraints and Modular Risk Allocation
标题：跨越共形预测和情景优化：抛弃的约束和模块化风险分配
链接：https://arxiv.org/abs/2603.19396
摘要：情景优化和保形预测有一个共同的目标，即把有限的样本转化为安全裕度。然而，不同的术语往往掩盖了各自保障之间的联系。本文直接从系统与控制的角度重新审视了这种联系。最近的保形/场景桥的\citet{OSullivanRomaoMargellos 2026}的基础上，我们扩展了可行的采样和丢弃场景算法的前进方向。具体来说，如果最终的决定是由一个稳定的子集所保留的采样约束，经典的平均违反法律承认一个直接的可交换性为基础的推导。在这种观点下，丢弃的样本自然会作为可接受的例外出现。我们还介绍了一个简单的模块化组成规则，将几个块校准证书组合成一个单一的联合保证。这条规则在多输出预测和有限时域控制中特别有用，因为工程师必须在坐标、约束或预测步骤中分配风险。最后，我们提供了一个校准的多步管周围的一个确定的预测数值说明。这些例子比较了不同阶段的风险分配，并强调了标准约束收紧问题中的性能和安全权衡。
摘要：Scenario optimization and conformal prediction share a common goal, that is, turning finite samples into safety margins. Yet, different terminology often obscures the connection between their respective guarantees. This paper revisits that connection directly from a systems-and-control viewpoint. Building on the recent conformal/scenario bridge of \citet{OSullivanRomaoMargellos2026}, we extend the forward direction to feasible sample-and-discard scenario algorithms. Specifically, if the final decision is determined by a stable subset of the retained sampled constraints, the classical mean violation law admits a direct exchangeability-based derivation. In this view, discarded samples naturally appear as admissible exceptions. We also introduce a simple modular composition rule that combines several blockwise calibration certificates into a single joint guarantee. This rule proves particularly useful in multi-output prediction and finite-horizon control, where engineers must distribute risk across coordinates, constraints, or prediction steps. Finally, we provide numerical illustrations using a calibrated multi-step tube around an identified predictor. These examples compare alternative stage-wise risk allocations and highlight the resulting performance and safety trade-offs in a standard constraint-tightening problem.
【5】Antenna Array Beamforming Based on a Hybrid Quantum Optimization Framework
标题：基于混合量子优化框架的天线阵束形成
链接：https://arxiv.org/abs/2603.20072
摘要：提出了一种离散相位和连续幅度联合优化的大规模天线阵列波束形成的混合量子优化框架。该方法结合量子启发搜索和经典的梯度细化，有效地处理混合离散-连续变量。在相位优化方面，采用格雷码和奇组合编码方案，提高了鲁棒性，避免了高阶伊辛模型的复杂性爆炸。对于幅度优化，几何自旋组合编码和两阶段的策略，使用量子启发的优化粗搜索和梯度优化精细细化。为了提高解决方案的多样性和质量，彩虹量子启发算法集成了多个优化器进行并行探索，然后进行基于层次聚类的候选细化。此外，提出了双外积方法和增广方法，有效地构造了耦合矩阵和偏置向量，提高了数值精度和实现效率。在第七届全国量子计算黑客竞赛评分规则下，对32元天线阵列的仿真结果表明，该方法在近主瓣旁瓣、宽角旁瓣、波束宽度和优化时间等约束条件下，获得了461.58分，比基线分数提高了近一倍。该框架为未来无线通信系统的波束形成优化提供了有效的参考。
摘要：This paper proposes a hybrid quantum optimization framework for large-scale antenna-array beamforming with jointly optimized discrete phases and continuous amplitudes. The method combines quantum-inspired search with classical gradient refinement to handle mixed discrete-continuous variables efficiently. For phase optimization, a Gray-code and odd-combination encoding scheme is introduced to improve robustness and avoid the complexity explosion of higher-order Ising models. For amplitude optimization, a geometric spin-combination encoding and a two-stage strategy are developed, using quantum-inspired optimization for coarse search and gradient optimization for fine refinement. To enhance solution diversity and quality, a rainbow quantum-inspired algorithm integrates multiple optimizers for parallel exploration, followed by hierarchical-clustering-based candidate refinement. In addition, a double outer-product method and an augmented version are proposed to construct the coupling matrix and bias vector efficiently, improving numerical precision and implementation efficiency. Under the scoring rules of the 7th National Quantum Computing Hackathon, simulations on a 32-element antenna array show that the proposed method achieves a score of 461.58 under constraints on near-main-lobe sidelobes, wide-angle sidelobes, beamwidth, and optimization time, nearly doubling the baseline score. The proposed framework provides an effective reference for beamforming optimization in future wireless communication systems.
【6】Infinite-dimensional spherical-radial decomposition for probabilistic functions, with application to constrained optimal control and Gaussian process regression
标题：概率函数的无限维球径向分解及其在约束最优控制和高斯过程回归中的应用
链接：https://arxiv.org/abs/2603.19907
备注：25 pages, 8 figures
摘要 ：球径向分解（SRD）是估计有限维椭圆分布上概率函数及其梯度的一种有效方法。在这项工作中，我们推广的SRD无限随机维结合子空间SRD与标准的蒙特卡罗方法。由此产生的方法，我们称之为混合无限维SRD（hiSRD）提供了一个无偏的，低方差的估计凸集产生的，例如，在机会约束优化。我们提供了一个理论分析的方差有限维SRD的维数增加，并表明，所提出的混合方法消除截断引起的偏见，减少方差，并允许概率函数的导数的计算。我们提出了全面的数值研究的风险中性随机PDE最优控制问题的联合机会状态约束，并优化核参数的高斯过程回归的约束下，后验过程满足联合机会约束。
摘要：The spherical-radial decomposition (SRD) is an efficient method for estimating probabilistic functions and their gradients defined over finite-dimensional elliptical distributions. In this work, we generalize the SRD to infinite stochastic dimensions by combining subspace SRD with standard Monte Carlo methods. The resulting method, which we call hybrid infinite-dimensional SRD (hiSRD) provides an unbiased, low-variance estimator for convex sets arising, for instance, in chance-constrained optimization. We provide a theoretical analysis of the variance of finite-dimensional SRD as the dimension increases, and show that the proposed hybrid method eliminates truncation-induced bias, reduces variance, and allows the computation of derivatives of probabilistic functions. We present comprehensive numerical studies for a risk-neutral stochastic PDE optimal control problem with joint chance state constraints, and for optimizing kernel parameters in Gaussian process regression under the constraint that the posterior process satisfies joint chance constraints.
预测|估计(9篇)
【1】Trojan horse hunt in deep forecasting models: Insights from the European Space Agency competition
标题：深度预测模型中的特洛伊木马搜寻：欧洲航天局竞赛的见解
链接：https://arxiv.org/abs/2603.20108
备注：43 pages, 18 figures
摘要：预测在现代安全关键应用中发挥着至关重要的作用，如空间操作。然而，越来越多地使用深度预测模型引入了特洛伊木马攻击的新安全风险，通过在训练数据中隐藏后门或直接在模型权重中进行。一旦植入，后门在测试时被特定的触发模式激活，导致模型产生操纵的预测。我们在我们的\textit{Trojan Horse Hunt}数据科学竞赛中关注这个问题，200多个团队面临着识别隐藏在航天器遥测深度预测模型中的触发器的任务。我们描述了新的任务制定，基准集，评估协议，并从竞争的最佳解决方案。我们进一步总结了有效识别时间序列预测模型中触发器的关键见解和研究方向。所有材料均可在官方竞赛网页https://www.kaggle.com/competitions/trojan-horse-hunt-in-space上公开获取。
摘要：Forecasting plays a crucial role in modern safety-critical applications, such as space operations. However, the increasing use of deep forecasting models introduces a new security risk of trojan horse attacks, carried out by hiding a backdoor in the training data or directly in the model weights. Once implanted, the backdoor is activated by a specific trigger pattern at test time, causing the model to produce manipulated predictions. We focus on this issue in our \textit{Trojan Horse Hunt} data science competition, where more than 200 teams faced the task of identifying triggers hidden in deep forecasting models for spacecraft telemetry. We describe the novel task formulation, benchmark set, evaluation protocol, and best solutions from the competition. We further summarize key insights and research directions for effective identification of triggers in time series forecasting models. All materials are publicly available on the official competition webpage https://www.kaggle.com/competitions/trojan-horse-hunt-in-space.
【2】HiPath: Hierarchical Vision-Language Alignment for Structured Pathology Report Prediction
标题：HiPath：用于结构化病理报告预测的分层视觉语言对齐
链接：https://arxiv.org/abs/2603.19957
备注：10 pages, 1 figures, 3 tables
摘要：病理学报告是结构化的多粒度文档，对一个或多个解剖部位的诊断结论、组织学等级和辅助测试结果进行编码;然而现有的病理学视觉语言模型（VLM）将此输出减少为平面标签或自由形式的文本。我们提出了HiPath，一个轻量级的VLM框架，建立在冻结的UNI 2和Qwen 3骨干上，将结构化报告预测作为其主要训练目标。三个可训练模块共1500万个参数解决了问题的互补方面：用于多图像视觉编码的分层补丁聚合器（HiPA），用于通过最佳传输进行跨模态对齐的分层对比学习（HiCL），以及用于结构化诊断生成的基于时隙的掩蔽诊断预测（Slot-MDP）。经过来自三家医院的749 K真实世界中国病理学病例的培训，HiPath实现了68.9%的严格准确率和74.7%的临床可接受准确率，以及97.3%的安全率，在相同的冷冻主干下优于所有基线。跨医院评估证实了泛化，严格准确性仅下降3.4pp，同时保持97.1%的安全性。
摘要：Pathology reports are structured, multi-granular documents encoding diagnostic conclusions, histological grades, and ancillary test results across one or more anatomical sites; yet existing pathology vision-language models (VLMs) reduce this output to a flat label or free-form text. We present HiPath, a lightweight VLM framework built on frozen UNI2 and Qwen3 backbones that treats structured report prediction as its primary training objective. Three trainable modules totalling 15M parameters address complementary aspects of the problem: a Hierarchical Patch Aggregator (HiPA) for multi-image visual encoding, Hierarchical Contrastive Learning (HiCL) for cross-modal alignment via optimal transport, and Slot-based Masked Diagnosis Prediction (Slot-MDP) for structured diagnosis generation. Trained on 749K real-world Chinese pathology cases from three hospitals, HiPath achieves 68.9% strict and 74.7% clinically acceptable accuracy with a 97.3% safety rate, outperforming all baselines under the same frozen backbone. Cross-hospital evaluation confirms generalisation with only a 3.4pp drop in strict accuracy while maintaining 97.1% safety.
【3】Eye Gaze-Informed and Context-Aware Pedestrian Trajectory Prediction in Shared Spaces with Automated Shuttles: A Virtual Reality Study
标题：使用自动航天飞机的共享空间中的眼神知情和上下文感知的行人轨迹预测：虚拟现实研究
链接：https://arxiv.org/abs/2603.19812
摘要：由于缺乏交通规则和复杂的行人互动，将自动穿梭车整合到共享的城市空间中面临着独特的挑战。因此，准确预测行人在这种非结构化环境中的行为对于确保安全和效率至关重要。本文介绍了一项虚拟现实（VR）研究，该研究捕捉了行人如何在不同的场景中与自动化穿梭机互动，包括不同的接近角度和在连续交通中导航。我们确定了行人在共享空间中的决策中存在的关键行为模式，包括犹豫、回避动作、凝视分配和情绪调整。为了对行人行为进行建模，我们提出了GazeX-LSTM，这是一种多模态的眼睛注视信息和上下文感知预测模型，它集成了行人的轨迹，细粒度的眼睛注视动态和上下文因素。我们通过利用眼动追踪数据来捕捉行人的注意力，将预测从以车辆为中心的角度转变为以人为中心的角度。我们系统地验证了独特的和不可替代的预测能力的眼睛注视单独的头部方向，进一步提高性能，通过整合上下文变量。值得注意的是，眼睛注视数据和上下文信息的组合产生了行人行为预测准确性的超加性改进，揭示了视觉注意力和情景上下文之间的互补关系。总之，我们的研究结果提供了第一个证据，即眼睛注视信息建模从根本上提高了行人行为预测，并突出了情境背景在共享空间交互中的关键作用。这为更安全、更具适应性的自动驾驶汽车技术铺平了道路，这些技术可以解释人们在复杂的共享空间中的感知和行为。
摘要 ：The integration of Automated Shuttles into shared urban spaces presents unique challenges due to the absence of traffic rules and the complex pedestrian interactions. Accurately anticipating pedestrian behavior in such unstructured environments is therefore critical for ensuring both safety and efficiency. This paper presents a Virtual Reality (VR) study that captures how pedestrians interact with automated shuttles across diverse scenarios, including varying approach angles and navigating in continuous traffic. We identify critical behavior patterns present in pedestrians' decision-making in shared spaces, including hesitation, evasive maneuvers, gaze allocation, and proxemic adjustments. To model pedestrian behavior, we propose GazeX-LSTM, a multimodal eye gaze-informed and context-aware prediction model that integrates pedestrians' trajectories, fine-grained eye gaze dynamics, and contextual factors. We shift prediction from a vehicle- to a human-centered perspective by leveraging eye-tracking data to capture pedestrian attention. We systematically validate the unique and irreplaceable predictive power of eye gaze over head orientation alone, further enhancing performance by integrating contextual variables. Notably, the combination of eye gaze data and contextual information produces super-additive improvements on pedestrian behavior prediction accuracy, revealing the complementary relationship between visual attention and situational contexts. Together, our findings provide the first evidence that eye gaze-informed modeling fundamentally advances pedestrian behavior prediction and highlight the critical role of situational contexts in shared-space interactions. This paves the way for safer and more adaptive automated vehicle technologies that account for how people perceive and act in complex shared spaces.
【4】Target Concept Tuning Improves Extreme Weather Forecasting
标题：目标概念调整改进极端天气预报
链接：https://arxiv.org/abs/2603.19325
摘要：用于气象预报的深度学习模型通常在罕见但影响大的事件中失败，例如台风，相关数据稀缺。现有的微调方法通常面临着忽略这些极端事件和过度拟合它们之间的权衡，以牺牲整体性能。我们提出了TaCT，一个可解释的概念门控微调框架，通过选择性的模型改进解决了上述问题：模型专门针对故障情况进行调整，同时在常见情况下保持性能。为此，TaCT使用稀疏自动编码器和反事实分析自动发现与故障相关的内部概念，并仅在相应概念被激活时更新参数，而不是应用统一的自适应。实验表明，在不降低其他气象变量的情况下，不同地区的台风预报得到了一致的改善。所确定的概念对应于物理上有意义的循环模式，揭示了模型偏差，并支持科学预测任务中值得信赖的适应。该代码可在https://anonymous.4open.science/r/Concept-Gated-Fine-tune-62AC上获得。
摘要：Deep learning models for meteorological forecasting often fail in rare but high-impact events such as typhoons, where relevant data is scarce. Existing fine-tuning methods typically face a trade-off between overlooking these extreme events and overfitting them at the expense of overall performance. We propose TaCT, an interpretable concept-gated fine-tuning framework that solves the aforementioned issue by selective model improvement: models are adapted specifically for failure cases while preserving performance in common scenarios. To this end, TaCT automatically discovers failure-related internal concepts using Sparse Autoencoders and counterfactual analysis, and updates parameters only when the corresponding concepts are activated, rather than applying uniform adaptation. Experiments show consistent improvements in typhoon forecasting across different regions without degrading other meteorological variables. The identified concepts correspond to physically meaningful circulation patterns, revealing model biases and supporting trustworthy adaptation in scientific forecasting tasks. The code is available at https://anonymous.4open.science/r/Concept-Gated-Fine-tune-62AC.
【5】LeWorldModel: Stable End-to-End Joint-Embedding Predictive Architecture from Pixels
标题：LeWorldModel：来自Pixels的稳定端到端关节嵌入预测架构
链接：https://arxiv.org/abs/2603.19312
摘要：联合嵌入预测架构（JEPAs）为在紧凑的潜在空间中学习世界模型提供了一个令人信服的框架，但现有的方法仍然很脆弱，依赖于复杂的多项损失，指数移动平均，预训练的编码器或辅助监督来避免表示崩溃。在这项工作中，我们介绍了LeWorldModel（LeWM），这是第一个JEPA，它只使用两个损失项从原始像素稳定地进行端到端训练：下一个嵌入预测损失和正则化器强制执行高斯分布的潜在嵌入。与现有的唯一端到端替代方案相比，这将可调损耗超参数从六个减少到一个。LeWM计划在几个小时内在单个GPU上训练约1500万个参数，比基于基础模型的世界模型快48倍，同时在各种2D和3D控制任务中保持竞争力。除了控制之外，我们还表明LeWM的潜在空间通过探测物理量编码了有意义的物理结构。意外评价证实，该模型可靠地检测到物理上不可信的事件。
摘要：Joint Embedding Predictive Architectures (JEPAs) offer a compelling framework for learning world models in compact latent spaces, yet existing methods remain fragile, relying on complex multi-term losses, exponential moving averages, pre-trained encoders, or auxiliary supervision to avoid representation collapse. In this work, we introduce LeWorldModel (LeWM), the first JEPA that trains stably end-to-end from raw pixels using only two loss terms: a next-embedding prediction loss and a regularizer enforcing Gaussian-distributed latent embeddings. This reduces tunable loss hyperparameters from six to one compared to the only existing end-to-end alternative. With ~15M parameters trainable on a single GPU in a few hours, LeWM plans up to 48x faster than foundation-model-based world models while remaining competitive across diverse 2D and 3D control tasks. Beyond control, we show that LeWM's latent space encodes meaningful physical structure through probing of physical quantities. Surprise evaluation confirms that the model reliably detects physically implausible events.
【6】A Dynamic Bayesian and Machine Learning Framework for Quantitative Evaluation and Prediction of Operator Situation Awareness in Nuclear Power Plants
标题：用于核电站操作员态势感知定量评估和预测的动态Bayesian和机器学习框架
链接：https://arxiv.org/abs/2603.19298
摘要：操作员态势感知是复杂核控制环境中人的可靠性的关键而难以捉摸的决定因素。现有的评估方法，如SAGAT和SART，仍然是静态的，回顾性的，并脱离不断发展的认知动态驱动操作风险。为了克服这些局限性，本研究引入了动态贝叶斯态势感知机器学习框架（DBML SA），这是一种统一的方法，融合了概率推理和数据驱动的智能，以实现定量，可解释和预测的态势感知建模。利用212个操作事件报告（2007年至2021年），该框架重建了跨多个认知层的11个绩效形成因素的因果时间结构。贝叶斯组件能够在不确定性下对态势感知可靠性进行时间演化推断，而神经组件建立了从PSF到SART分数的非线性预测映射，实现了平均绝对百分比误差为13.8%，与主观评价具有统计学一致性（p > 0.05）。结果强调，培训质量和压力动态的主要驱动力的态势感知退化。总的来说，DBML SA通过实现实时认知监控、敏感性分析和预警预测，超越了传统的基于数据库的评估，为下一代数字主控制室的智能人机可靠性管理铺平了道路。
摘要：Operator situation awareness is a pivotal yet elusive determinant of human reliability in complex nuclear control environments. Existing assessment methods, such as SAGAT and SART, remain static, retrospective, and detached from the evolving cognitive dynamics that drive operational risk. To overcome these limitations, this study introduces the dynamic Bayesian machine learning framework for situation awareness (DBML SA), a unified approach that fuses probabilistic reasoning and data driven intelligence to achieve quantitative, interpretable, and predictive situation awareness modeling. Leveraging 212 operational event reports (2007 to 2021), the framework reconstructs the causal temporal structure of 11 performance shaping factors across multiple cognitive layers. The Bayesian component enables time evolving inference of situation awareness reliability under uncertainty, while the neural component establishes a nonlinear predictive mapping from PSFs to SART scores, achieving a mean absolute percentage error of 13.8 % with statistical consistency to subjective evaluations (p > 0.05). Results highlight training quality and stress dynamics as primary drivers of situation awareness degradation. Overall, DBML SA transcends traditional questionnaire-based assessments by enabling real-time cognitive monitoring, sensitivity analysis, and early-warning prediction, paving the way toward intelligent human machine reliability management in next-generation digital main control rooms.
【7】Deep Autocorrelation Modeling for Time-Series Forecasting: Progress and Prospects
标题：时间序列预测的深度自相关建模：进展与前景
链接：https://arxiv.org/abs/2603.19899
摘要：自相关性是时间序列数据的一个定义特征，其中每个观测值在统计上依赖于其先前观测值。在深度时间序列预测的背景下，自相关出现在输入历史和标签序列中，提出了两个主要的研究挑战：（1）设计神经架构，模拟历史序列中的自相关，以及（2）设计学习目标，模拟标签序列中的自相关。最近的研究在应对这些挑战方面取得了进展，但仍然缺乏对这两个方面的系统调查。为了弥补这一差距，本文从自相关建模的角度对深度时间序列预测进行了全面的回顾。与现有的调查相比，这项工作有两个独特的贡献。首先，它提出了一个新的分类法，包括最近的文献模型架构和学习目标-而以前的调查忽视或不充分讨论后一方面。其次，它从统一的、以自相关为中心的角度对调查文献的动机、见解和进展进行了全面分析，全面概述了深度时间序列预测的演变。论文和资源的完整列表可在https://github.com/Master-PLC/Awesome-TSF-Papers上获得。
摘要：Autocorrelation is a defining characteristic of time-series data, where each observation is statistically dependent on its predecessors. In the context of deep time-series forecasting, autocorrelation arises in both the input history and the label sequences, presenting two central research challenges: (1) designing neural architectures that model autocorrelation in history sequences, and (2) devising learning objectives that model autocorrelation in label sequences. Recent studies have made strides in tackling these challenges, but a systematic survey examining both aspects remains lacking. To bridge this gap, this paper provides a comprehensive review of deep time-series forecasting from the perspective of autocorrelation modeling. In contrast to existing surveys, this work makes two distinctive contributions. First, it proposes a novel taxonomy that encompasses recent literature on both model architectures and learning objectives -- whereas prior surveys neglect or inadequately discuss the latter aspect. Second, it offers a thorough analysis of the motivations, insights, and progression of the surveyed literature from a unified, autocorrelation-centric perspective, providing a holistic overview of the evolution of deep time-series forecasting. The full list of papers and resources is available at https://github.com/Master-PLC/Awesome-TSF-Papers.
【8】Model Selection and Parameter Estimation of Multi-dimensional Gaussian Mixture Model
标题：多维高斯混合模型的模型选择和参数估计
链接：https://arxiv.org/abs/2603.19657
摘要：本文研究了多维高斯混合模型的学习问题，重点研究了模型阶数的选择和混合分布的有效估计。我们首先建立可靠模型选择所需的临界样本复杂度的信息论下限。更具体地说，我们表明，从一个简单的模型区分$k$组分混合物需要一个样本大小缩放$Ω（Δ^{-（4k-4）}）$。然后，我们提出了一个基于阈值的估计算法，评估从随机傅立叶测量向量构造的经验协方差矩阵的频谱间隙。这种无参数估计器的有效时间复杂度为$\mathcal{O}（k^2 n）$，随样本大小线性缩放。我们证明了我们的方法的样本复杂度与所建立的下限相匹配，证实了其关于组件分离距离$Δ$的最小最大最优性。 在估计模型阶数的条件下，我们随后介绍了一种基于梯度的最小化方法进行参数估计。为了有效地导航非凸目标景观，我们采用了数据驱动的，基于分数的初始化策略，保证快速收敛。我们证明了该方法达到了估计分量均值的最优参数收敛速度$\mathcal{O}_p（n^{-1/2}）$。为了提高算法在环境维度超过混合成分数量的高维区域中的效率（即，\（d > k\）），结合主成分分析（PCA）进行降维。数值实验表明，我们的基于傅立叶的算法框架优于传统的期望最大化（EM）方法在估计精度和计算时间。
摘要：In this paper, we study the problem of learning multi-dimensional Gaussian Mixture Models (GMMs), with a specific focus on model order selection and efficient mixing distribution estimation. We first establish an information-theoretic lower bound on the critical sample complexity required for reliable model selection. More specifically, we show that distinguishing a $k$-component mixture from a simpler model necessitates a sample size scaling of $Ω(Δ^{-(4k-4)})$. We then propose a thresholding-based estimation algorithm that evaluates the spectral gap of an empirical covariance matrix constructed from random Fourier measurement vectors. This parameter-free estimator operates with an efficient time complexity of $\mathcal{O}(k^2 n)$, scaling linearly with the sample size. We demonstrate that the sample complexity of our method matches the established lower bound, confirming its minimax optimality with respect to the component separation distance $Δ$. Conditioned on the estimated model order, we subsequently introduce a gradient-based minimization method for parameter estimation. To effectively navigate the non-convex objective landscape, we employ a data-driven, score-based initialization strategy that guarantees rapid convergence. We prove that this method achieves the optimal parametric convergence rate of $\mathcal{O}_p(n^{-1/2})$ for estimating the component means. To enhance the algorithm's efficiency in high-dimensional regimes where the ambient dimension exceeds the number of mixture components (i.e., \(d > k\)), we integrate principal component analysis (PCA) for dimension reduction. Numerical experiments demonstrate that our Fourier-based algorithmic framework outperforms conventional Expectation-Maximization (EM) methods in both estimation accuracy and computational time.
【9】Generalized Stock Price Prediction for Multiple Stocks Combined with News Fusion
标题：结合新闻融合的多只股票广义股价预测
链接：https://arxiv.org/abs/2603.19286
备注：Accepted to Journal of Information Science and Engineering (JISE)
摘要：Predicting stock prices presents challenges in financial forecasting. While traditional approaches such as ARIMA and RNNs are prevalent, recent developments in Large Language Models (LLMs) offer alternative methodologies. This paper introduces an approach that integrates LLMs with daily financial news for stock price prediction. To address the challenge of processing news data and identifying relevant content, we utilize stock name embeddings within attention mechanisms. Specifically, we encode news articles using a pre-trained LLM and implement three attention-based pooling techniques -- self-attentive, cross-attentive, and position-aware self-attentive pooling -- to filter news based on stock relevance. The filtered news embeddings, combined with historical stock prices, serve as inputs to the prediction model. Unlike prior studies that focus on individual stocks, our method trains a single generalized model applicable across multiple stocks. Experimental results demonstrate a 7.11% reduction in Mean Absolute Error (MAE) compared to the baseline, indicating the utility of stock name embeddings for news filtering and price forecasting within a generalized framework.
其他神经网络|深度学习|模型|建模(25篇)
【1】Kolmogorov-Arnold causal generative models
标题：Kolmogorov-Arnold因果生成模型
链接：https://arxiv.org/abs/2603.20184
备注：14 pages, 8 figures, 3 tables, 5 algorithms, preprint
摘要 ：Causal generative models provide a principled framework for answering observational, interventional, and counterfactual queries from observational data. However, many deep causal models rely on highly expressive architectures with opaque mechanisms, limiting auditability in high-stakes domains. We propose KaCGM, a causal generative model for mixed-type tabular data where each structural equation is parameterized by a Kolmogorov--Arnold Network (KAN). This decomposition enables direct inspection of learned causal mechanisms, including symbolic approximations and visualization of parent--child relationships, while preserving query-agnostic generative semantics. We introduce a validation pipeline based on distributional matching and independence diagnostics of inferred exogenous variables, allowing assessment using observational data alone. Experiments on synthetic and semi-synthetic benchmarks show competitive performance against state-of-the-art methods. A real-world cardiovascular case study further demonstrates the extraction of simplified structural equations and interpretable causal effects. These results suggest that expressive causal generative modeling and functional transparency can be achieved jointly, supporting trustworthy deployment in tabular decision-making settings. Code: https://github.com/aalmodovares/kacgm
【2】How Out-of-Equilibrium Phase Transitions can Seed Pattern Formation in Trained Diffusion Models
标题：非平衡相转变如何在经过训练的扩散模型中促成模式的形成
链接：https://arxiv.org/abs/2603.20092
摘要：In this work, we propose a theoretical framework that interprets the generation process in trained diffusion models as an instance of out-of-equilibrium phase transitions. We argue that, rather than evolving smoothly from noise to data, reverse diffusion passes through a critical regime in which small spatial fluctuations are amplified and seed the emergence of large-scale structure. Our central insight is that architectural constraints, such as locality, sparsity, and translation equivariance, transform memorization-driven instabilities into collective spatial modes, enabling the formation of coherent patterns beyond the training data. Using analytically tractable patch score models, we show how classical symmetry-breaking bifurcations generalize into spatially extended critical phenomena described by softening Fourier modes and growing correlation lengths. We further connect these dynamics to effective field theories of the Ginzburg-Landau type and to mechanisms of pattern formation in non-equilibrium physics. Empirical results on trained convolutional diffusion models corroborate the theory, revealing signatures of criticality including mode softening and rapid growth of spatial correlations. Finally, we demonstrate that this critical regime has practical relevance: targeted perturbations, such as classifier-free guidance pulses applied at the estimated critical time, significantly improve generation control. Together, these findings position non-equilibrium critical phenomena as a unifying principle for understanding, and potentially improving, the behavior of modern diffusion models.
【3】Continual Learning as Shared-Manifold Continuation Under Compatible Shift
标题：兼容轮班下作为共享Manifold延续的连续学习
链接：https://arxiv.org/abs/2603.20036
备注：11 pages, 4 figures, repo: https://github.com/kkryon/spma
摘要：Continual learning methods usually preserve old behavior by regularizing parameters, matching old outputs, or replaying previous examples. These strategies can reduce forgetting, but they do not directly specify how the latent representation should evolve. We study a narrower geometric alternative for the regime where old and new data should remain on the same latent support: continual learning as continuation of a shared manifold. We instantiate this view within Support-Preserving Manifold Assimilation (SPMA) and evaluate a geometry-preserving variant, SPMA-OG, that combines sparse replay, output distillation, relational geometry preservation, local smoothing, and chart-assignment regularization on old anchors. On representative compatible-shift CIFAR10 and Tiny-ImageNet runs, SPMA-OG improves over sparse replay baselines in old-task retention and representation-preservation metrics while remaining competitive on new-task accuracy. On a controlled synthetic atlas-manifold benchmark, it achieves near-perfect anchor-geometry preservation while also improving new-task accuracy over replay. These results provide evidence that geometry-aware anchor regularization is a useful inductive bias when continual learning should preserve a shared latent support rather than create a new one.
【4】Model-Driven Learning-Based Physical Layer Authentication for Mobile Wi-Fi Devices
标题：移动Wi-Fi设备的模型驱动的基于学习的物理层认证
链接：https://arxiv.org/abs/2603.19972
摘要：The rise of wireless technologies has made the Internet of Things (IoT) ubiquitous, but the broadcast nature of wireless communications exposes IoT to authentication risks. Physical layer authentication (PLA) offers a promising solution by leveraging unique characteristics of wireless channels. As a common approach in PLA, hypothesis testing yields a theoretically optimal Neyman-Pearson (NP) detector, but its reliance on channel statistics limits its practicality in real-world scenarios. In contrast, deep learning-based PLA approaches are practical but tend to be not optimal. To address these challenges, we proposed a learning-based PLA scheme driven by hypothesis testing and conducted extensive simulations and experimental evaluations using Wi-Fi. Specifically, we incorporated conditional statistical models into the hypothesis testing framework to derive a theoretically optimal NP detector. Building on this, we developed LiteNP-Net, a lightweight neural network driven by the NP detector. Simulation results demonstrated that LiteNP-Net could approach the performance of the NP detector even without prior knowledge of the channel statistics. To further assess its effectiveness in practical environments, we deployed an experimental testbed using Wi-Fi IoT development kits in various real-world scenarios. Experimental results demonstrated that the LiteNP-Net outperformed the conventional correlation-based method as well as state-of-the-art Siamese-based methods.
【5】NASimJax: GPU-Accelerated Policy Learning Framework for Penetration Testing
标题：NASimJax：用于渗透测试的GOP加速政策学习框架
链接：https://arxiv.org/abs/2603.19864
摘要 ：Penetration testing, the practice of simulating cyberattacks to identify vulnerabilities, is a complex sequential decision-making task that is inherently partially observable and features large action spaces. Training reinforcement learning (RL) policies for this domain faces a fundamental bottleneck: existing simulators are too slow to train on realistic network scenarios at scale, resulting in policies that fail to generalize. We present NASimJax, a complete JAX-based reimplementation of the Network Attack Simulator (NASim), achieving up to 100x higher environment throughput than the original simulator. By running the entire training pipeline on hardware accelerators, NASimJax enables experimentation on larger networks under fixed compute budgets that were previously infeasible. We formulate automated penetration testing as a Contextual POMDP and introduce a network generation pipeline that produces structurally diverse and guaranteed-solvable scenarios. Together, these provide a principled basis for studying zero-shot policy generalization. We use the framework to investigate action-space scaling and generalization across networks of up to 40 hosts. We find that Prioritized Level Replay better handles dense training distributions than Domain Randomization, particularly at larger scales, and that training on sparser topologies yields an implicit curriculum that improves out-of-distribution generalization, even on topologies denser than those seen during training. To handle linearly growing action spaces, we propose a two-stage action decomposition (2SAS) that substantially outperforms flat action masking at scale. Finally, we identify a failure mode arising from the interaction between Prioritized Level Replay's episode-reset behaviour and 2SAS's credit assignment structure. NASimJax thus provides a fast, flexible, and realistic platform for advancing RL-based penetration testing.
【6】Failure Modes for Deep Learning-Based Online Mapping: How to Measure and Address Them
标题：基于深度学习的在线地图的失败模式：如何测量和解决它们
链接：https://arxiv.org/abs/2603.19852
备注：Accepted to CVPR 2026, final camera ready version is published there
摘要：Deep learning-based online mapping has emerged as a cornerstone of autonomous driving, yet these models frequently fail to generalize beyond familiar environments. We propose a framework to identify and measure the underlying failure modes by disentangling two effects: Memorization of input features and overfitting to known map geometries. We propose measures based on evaluation subsets that control for geographical proximity and geometric similarity between training and validation scenes. We introduce Fréchet distance-based reconstruction statistics that capture per-element shape fidelity without threshold tuning, and define complementary failure-mode scores: a localization overfitting score quantifying the performance drop when geographic cues disappear, and a map geometry overfitting score measuring degradation as scenes become geometrically novel. Beyond models, we analyze dataset biases and contribute map geometry-aware diagnostics: A minimum-spanning-tree (MST) diversity measure for training sets and a symmetric coverage measure to quantify geometric similarity between splits. Leveraging these, we formulate an MST-based sparsification strategy that reduces redundancy and improves balancing and performance while shrinking training size. Experiments on nuScenes and Argoverse 2 across multiple state-of-the-art models yield more trustworthy assessment of generalization and show that map geometry-diverse and balanced training sets lead to improved performance. Our results motivate failure-mode-aware protocols and map geometry-centric dataset design for deployable online mapping.
【7】Two-Time-Scale Learning Dynamics: A Population View of Neural Network Training
标题：两时间尺度学习动力学：神经网络训练的总体观点
链接：https://arxiv.org/abs/2603.19808
摘要：Population-based learning paradigms, including evolutionary strategies, Population-Based Training (PBT), and recent model-merging methods, combine fast within-model optimisation with slower population-level adaptation. Despite their empirical success, a general mathematical description of the resulting collective training dynamics remains incomplete. We introduce a theoretical framework for neural network training based on two-time-scale population dynamics. We model a population of neural networks as an interacting agent system in which network parameters evolve through fast noisy gradient updates of SGD/Langevin type, while hyperparameters evolve through slower selection--mutation dynamics. We prove the large-population limit for the joint distribution of parameters and hyperparameters and, under strong time-scale separation, derive a selection--mutation equation for the hyperparameter density. For each fixed hyperparameter, the fast parameter dynamics relaxes to a Boltzmann--Gibbs measure, inducing an effective fitness for the slow evolution. The averaged dynamics connects population-based learning with bilevel optimisation and classical replicator--mutator models, yields conditions under which the population mean moves toward the fittest hyperparameter, and clarifies the role of noise and diversity in balancing optimisation and exploration. Numerical experiments illustrate both the large-population regime and the reduced two-time-scale dynamics, and indicate that access to the effective fitness, either in closed form or through population-level estimation, can improve population-level updates.
【8】Scalable Learning of Multivariate Distributions via Coresets
标题：通过核心集进行多元分布的可扩展学习
链接：https://arxiv.org/abs/2603.19792
备注：AISTATS 2026
摘要：Efficient and scalable non-parametric or semi-parametric regression analysis and density estimation are of crucial importance to the fields of statistics and machine learning. However, available methods are limited in their ability to handle large-scale data. We address this issue by developing a novel coreset construction for multivariate conditional transformation models (MCTMs) to enhance their scalability and training efficiency. To the best of our knowledge, these are the first coresets for semi-parametric distributional models. Our approach yields substantial data reduction via importance sampling. It ensures with high probability that the log-likelihood remains within multiplicative error bounds of $(1\pm\varepsilon)$ and thereby maintains statistical model accuracy. Compared to conventional full-parametric models, where coresets have been incorporated before, our semi-parametric approach exhibits enhanced adaptability, particularly in scenarios where complex distributions and non-linear relationships are present, but not fully understood. To address numerical problems associated with normalizing logarithmic terms, we follow a geometric approximation based on the convex hull of input data. This ensures feasible, stable, and accurate inference in scenarios involving large amounts of data. Numerical experiments demonstrate substantially improved computational efficiency when handling large and complex datasets, thus laying the foundation for a broad range of applications within the statistics and machine learning communities.
【9】Growing Networks with Autonomous Pruning
标题：通过自主修剪来发展网络
链接：https://arxiv.org/abs/2603.19759
摘要：This paper introduces Growing Networks with Autonomous Pruning (GNAP) for image classification. Unlike traditional convolutional neural networks, GNAP change their size, as well as the number of parameters they are using, during training, in order to best fit the data while trying to use as few parameters as possible. This is achieved through two complementary mechanisms: growth and pruning. GNAP start with few parameters, but their size is expanded periodically during training to add more expressive power each time the network has converged to a saturation point. Between these growing phases, model parameters are trained for classification and pruned simultaneously, with complete autonomy by gradient descent. Growing phases allow GNAP to improve their classification performance, while autonomous pruning allows them to keep as few parameters as possible. Experimental results on several image classification benchmarks show that our approach can train extremely sparse neural networks with high accuracy. For example, on MNIST, we achieved 99.44% accuracy with as few as 6.2k parameters, while on CIFAR10, we achieved 92.2\ accuracy with 157.8k parameters.
【10】Learning from Similarity/Dissimilarity and Pairwise Comparison
标题：从相似性/不同性和成对比较中学习
链接：https://arxiv.org/abs/2603.19713
摘要：This paper addresses binary classification in scenarios where obtaining explicit instance level labels is impractical, by exploiting multiple weak labels defined on instance pairs. The existing SconfConfDiff classification framework relies on continuous valued probabilistic supervision, including similarity-confidence, the probability of class agreement, and confidence-difference, the difference in positive class probabilities. However, probabilistic labeling requires subjective uncertainty quantification, often leading to unstable supervision. We propose SD-Pcomp classification, a binary judgment based weakly supervised learning framework that relies only on relative judgments, namely class agreement between two instances and pairwise preference toward the positive class. The method employs Similarity/Dissimilarity (SD) labels and Pairwise Comparison (Pcomp) labels, and develops two unbiased risk estimators, (i) a convex combination of SD and Pcomp and (ii) a unified estimator that integrates both labels by modeling their relationship. Theoretical analysis and experimental results show that the proposed approach improves classification performance over methods using a single weak label, and is robust to label noise and uncertainty in class prior estimation.
【11】Diminishing Returns in Expanding Generative Models and Godel-Tarski-Lob Limits
标题：扩展生成模型中的回报递减和Godel-Tarski-Lob极限
链接：https://arxiv.org/abs/2603.19687
摘要：Modern generative modelling systems are increasingly improved by expanding model capacity, training data, and computational resources. While empirical studies have documented such scaling behaviour across architectures including generative adversarial networks, variational autoencoders, transformer-based models, and diffusion models, the theoretical limits of capability growth in expanding generative systems remain poorly understood. In this paper we develop a general task-space framework for analysing expanding generative reasoning systems. Each system induces a subset of a global task space representing the tasks it can successfully solve, and system capability is measured by the probability mass of this solved-task set under a fixed task distribution. Within this framework we prove a structural result showing that, under mild assumptions, the marginal improvement in solved tasks must converge to zero as system capacity increases. Thus expanding generative systems may continue to gain capability, but the probability mass of newly solvable tasks necessarily diminishes asymptotically. We further provide a prediction-theoretic refinement based on complexity-weighted hypothesis classes inspired by algorithmic probability, yielding quantitative bounds on marginal improvement in prediction settings. Finally, we examine logical reasoning tasks and show that classical results from mathematical logic -- including Rosser incompleteness, Tarski's undefinability theorem, and Löb's theorem -- imply the persistence of unresolved logical tasks within sufficiently expressive reasoning systems. Together these results provide a mathematical perspective on the asymptotic behaviour of expanding generative systems, showing that long-run capability growth is constrained both by diminishing marginal improvements in task coverage and by fundamental logical limitations on internal reasoning.
【12】Wearable Foundation Models Should Go Beyond Static Encoders
标题：可穿戴基础型号应超越静态编码器
链接：https://arxiv.org/abs/2603.19564
备注：13 pages
摘要：Wearable foundation models (WFMs), trained on large volumes of data collected by affordable, always-on devices, have demonstrated strong performance on short-term, well-defined health monitoring tasks, including activity recognition, fitness tracking, and cardiovascular signal assessment. However, most existing WFMs primarily map short temporal windows to predefined labels via static encoders, emphasizing retrospective prediction rather than reasoning over evolving personal history, context, and future risk trajectories. As a result, they are poorly suited for modeling chronic, progressive, or episodic health conditions that unfold over weeks, months or years. Hence, we argue that WFMs must move beyond static encoders and be explicitly designed for longitudinal, anticipatory health reasoning. We identify three foundational shifts required to enable this transition: (1) Structurally rich data, which goes beyond isolated datasets or outcome-conditioned collection to integrated multimodal, long-term personal trajectories, and contextual metadata, ideally supported by open and interoperable data ecosystems; (2) Longitudinal-aware multimodal modeling, which prioritizes long-context inference, temporal abstraction, and personalization over cross-sectional or population-level prediction; and (3) Agentic inference systems, which move beyond static prediction to support planning, decision-making, and clinically grounded intervention under uncertainty. Together, these shifts reframe wearable health monitoring from retrospective signal interpretation toward continuous, anticipatory, and human-aligned health support.
【13】Subspace Kernel Learning on Tensor Sequences
标题：张量序列的子空间核学习
链接：https://arxiv.org/abs/2603.19546
备注：Accepted at the Fourteenth International Conference on Learning Representations (ICLR 2026)
摘要：Learning from structured multi-way data, represented as higher-order tensors, requires capturing complex interactions across tensor modes while remaining computationally efficient. We introduce Uncertainty-driven Kernel Tensor Learning (UKTL), a novel kernel framework for $M$-mode tensors that compares mode-wise subspaces derived from tensor unfoldings, enabling expressive and robust similarity measure. To handle large-scale tensor data, we propose a scalable Nyström kernel linearization with dynamically learned pivot tensors obtained via soft $k$-means clustering. A key innovation of UKTL is its uncertainty-aware subspace weighting, which adaptively down-weights unreliable mode components based on estimated confidence, improving robustness and interpretability in comparisons between input and pivot tensors. Our framework is fully end-to-end trainable and naturally incorporates both multi-way and multi-mode interactions through structured kernel compositions. Extensive evaluations on action recognition benchmarks (NTU-60, NTU-120, Kinetics-Skeleton) show that UKTL achieves state-of-the-art performance, superior generalization, and meaningful mode-wise insights. This work establishes a principled, scalable, and interpretable kernel learning paradigm for structured multi-way and multi-modal tensor sequences.
【14】Verifiable Error Bounds for Physics-Informed Neural Network Solutions of Lyapunov and Hamilton-Jacobi-Bellman Equations
标题：基于物理信息的神经网络求解Lyapunov和Hamilton-Jacobi-Bellman方程的可验证误差界
链接：https://arxiv.org/abs/2603.19545
摘要：Many core problems in nonlinear systems analysis and control can be recast as solving partial differential equations (PDEs) such as Lyapunov and Hamilton-Jacobi-Bellman (HJB) equations. Physics-informed neural networks (PINNs) have emerged as a promising mesh-free approach for approximating their solutions, but in most existing works there is no rigorous guarantee that a small PDE residual implies a small solution error. This paper develops verifiable error bounds for approximate solutions of Lyapunov and HJB equations, with particular emphasis on PINN-based approximations. For both the Lyapunov and HJB PDEs, we show that a verifiable residual bound yields relative error bounds with respect to the true solutions as well as computable a posteriori estimates in terms of the approximate solutions. For the HJB equation, this also yields certified upper and lower bounds on the optimal value function on compact sublevel sets and quantifies the optimality gap of the induced feedback policy. We further show that one-sided residual bounds already imply that the approximation itself defines a valid Lyapunov or control Lyapunov function. We illustrate the results with numerical examples.
【15】Any-Subgroup Equivariant Networks via Symmetry Breaking
标题：通过对称性破缺的任意子集等变网络
链接：https://arxiv.org/abs/2603.19486
备注：Accepted at ICLR 2026
摘要：The inclusion of symmetries as an inductive bias, known as equivariance, often improves generalization on geometric data (e.g. grids, sets, and graphs). However, equivariant architectures are usually highly constrained, designed for symmetries chosen a priori, and not applicable to datasets with other symmetries. This precludes the development of flexible, multi-modal foundation models capable of processing diverse data equivariantly. In this work, we build a single model -- the Any-Subgroup Equivariant Network (ASEN) -- that can be simultaneously equivariant to several groups, simply by modulating a certain auxiliary input feature. In particular, we start with a fully permutation-equivariant base model, and then obtain subgroup equivariance by using a symmetry-breaking input whose automorphism group is that subgroup. However, finding an input with the desired automorphism group is computationally hard. We overcome this by relaxing from exact to approximate symmetry breaking, leveraging the notion of 2-closure to derive fast algorithms. Theoretically, we show that our subgroup-equivariant networks can simulate equivariant MLPs, and their universality can be guaranteed if the base model is universal. Empirically, we validate our method on symmetry selection for graph and image tasks, as well as multitask and transfer learning for sequence tasks, showing that a single network equivariant to multiple permutation subgroups outperforms both separate equivariant models and a single non-equivariant model.
【16】Do Post-Training Algorithms Actually Differ? A Controlled Study Across Model Scales Uncovers Scale-Dependent Ranking Inversions
标题：训练后算法真的不同吗？跨模型量表的对照研究揭示了量表依赖的排名倒置
链接：https://arxiv.org/abs/2603.19335
摘要：Post-training alignment has produced dozens of competing algorithms -- DPO, SimPO, KTO, GRPO, and others -- yet practitioners lack controlled comparisons to guide algorithm selection. We present OXRL, a unified framework implementing 51 post-training algorithms with identical infrastructure, enabling the first large-scale apples-to-apples evaluation. Our study spans 8 algorithms across 4 model scales (0.5B--7B), 3 evaluation domains, and a 20-variant DPO taxonomy (100 runs at 1.5B, 5 seeds each), totaling $\sim$240 training runs on H100 GPUs. Three headline findings emerge. (1)~Algorithm rankings are unstable across scale: at 1.5B, online RL (SGRPO) tops all methods at 58.0\%~$\pm$0.57 on GSM8K; by 7B, the worst small-scale method (SimPO) becomes the best (85.8\%), a complete ranking inversion driven by model scale rather than LoRA regularization (confirmed via 2$\times$2 factorial). (2)~Loss function modifications yield negligible gains: none of 20 DPO variants significantly outperform vanilla DPO after Bonferroni correction; the sole significant outlier, SimPO, is worse ($-$11.5~pp, $p < 10^{-4}$). (3)~Algorithm leverage is task-specific: the 19.3~pp GSM8K spread collapses to 0.54~pp on MATH ($36\times$) and 0.47~pp on general-domain benchmarks ($41\times$), confirming that algorithm choice matters primarily within the training distribution. These findings yield a hierarchy of leverage for practitioners: model scale (${\sim}$50~pp) $\gg$ training paradigm (${\sim}$10~pp) $\gg$ online vs.\ offline (${\sim}$9~pp) $\gg$ loss function (${\sim}$1~pp). We release all code, configs, and evaluation data as a living community benchmark.
【17】A General Deep Learning Framework for Wireless Resource Allocation under Discrete Constraints
标题：离散约束下无线资源分配的通用深度学习框架
链接：https://arxiv.org/abs/2603.19322
摘要：While deep learning (DL)-based methods have achieved remarkable success in continuous wireless resource allocation, efficient solutions for problems involving discrete variables remain challenging. This is primarily due to the zero-gradient issue in backpropagation, the difficulty of enforcing intricate constraints with discrete variables, and the inability in generating solutions with non-same-parameter-same-decision (non-SPSD) property. To address these challenges, this paper proposes a general DL framework by introducing the support set to represent the discrete variables. We model the elements of the support set as random variables and learn their joint probability distribution. By factorizing the joint probability as the product of conditional probabilities, each conditional probability is sequentially learned. This probabilistic modeling directly tackles all the aforementioned challenges of DL for handling discrete variables. By operating on probability distributions instead of hard binary decisions, the framework naturally avoids the zero-gradient issue. During the learning of the conditional probabilities, discrete constraints can be seamlessly enforced by masking out infeasible solutions. Moreover, with a dynamic context embedding that captures the evolving discrete solutions, the non-SPSD property is inherently provided by the proposed framework. We apply the proposed framework to two representative mixed-discrete wireless resource allocation problems: (a) joint user association and beamforming in cell-free systems, and (b) joint antenna positioning and beamforming in movable antenna-aided systems. Simulation results demonstrate that the proposed DL framework consistently outperforms existing baselines in terms of both system performance and computational efficiency.
【18】Towards Solving Polynomial-Objective Integer Programming with Hypergraph Neural Networks
标题：利用超图神经网络解决多项目标子索规划
链接：https://arxiv.org/abs/2603.19318
备注：Accepted for publication in CPAIOR 2026, to appear in Springer Lecture Notes in Computer Science (LNCS)
摘要：Complex real-world optimization problems often involve both discrete decisions and nonlinear relationships between variables. Many such problems can be modeled as polynomial-objective integer programs, encompassing cases with quadratic and higher-degree variable interactions. Nonlinearity makes them more challenging than their linear counterparts. In this paper, we propose a hypergraph neural network (HNN) based method to solve polynomial-objective integer programming (POIP). Besides presenting a high-degree-term-aware hypergraph representation to capture both high-degree information and variable-constraint interdependencies, we also propose a hypergraph neural network, which integrates convolution between variables and high-degree terms alongside convolution between variables and constraints, to predict solution values. Finally, a search process initialized from the predicted solutions is performed to further refine the results. Comprehensive experiments across a range of benchmarks demonstrate that our method consistently outperforms both existing learning-based approaches and state-of-the-art solvers, delivering superior solution quality with favorable efficiency. Note that our experiments involve both polynomial objectives and constraints, demonstrating our HNN's versatility for general POIP problems and highlighting its advancement over the existing literature.
【19】BrainSCL: Subtype-Guided Contrastive Learning for Brain Disorder Diagnosis
标题：BrainSCS：用于脑部疾病诊断的亚型引导对比学习
链接：https://arxiv.org/abs/2603.19295
摘要：Mental disorder populations exhibit pronounced heterogeneity -- that is, the significant differences between samples -- poses a significant challenge to the definition of positive pairs in contrastive learning. To address this, we propose a subtype-guided contrastive learning framework that models patient heterogeneity as latent subtypes and incorporates them as structural priors to guide discriminative representation learning. Specifically, we construct multi-view representations by combining patients' clinical text with graph structure adaptively learned from BOLD signals, to uncover latent subtypes via unsupervised spectral clustering. A dual-level attention mechanism is proposed to construct prototypes for capturing stable subtype-specific connectivity patterns. We further propose a subtype-guided contrastive learning strategy that pulls samples toward their subtype prototype graph, reinforcing intra-subtype consistency for providing effective supervisory signals to improve model performance. We evaluate our method on Major Depressive Disorder (MDD), Bipolar Disorder (BD), and Autism Spectrum Disorders (ASD). Experimental results confirm the effectiveness of subtype prototype graphs in guiding contrastive learning and demonstrate that the proposed approach outperforms state-of-the-art approaches. Our code is available at https://anonymous.4open.science/r/BrainSCL-06D7.
【20】Beam-aware Kernelized Contextual Bandits for User Association and Beamforming in mmWave Vehicular Networks
标题：毫米波车载网络中用于用户关联和束成形的束感知核化上下文带宽
链接：https://arxiv.org/abs/2603.19285
摘要：Timely channel information is necessary for vehicles to determine both the serving base station (BS) and the beamforming vector, but frequent estimation of fast-fading mmWave channels incurs significant overhead. To address this challenge, we propose a Beam-aware Kernelized Contextual Upper Confidence Bound (BKC-UCB) algorithm that estimates instantaneous transmission rates without additional channel measurements by exploiting historical contexts such as vehicle location and velocity, together with past observed transmission rates. Specifically, BKC-UCB leverages kernel methods to capture the nonlinear relationship between context and transmission rate by mapping contexts into a reproducing kernel Hilbert space (RKHS), where linear learning becomes feasible. Rather than treating each beam as an independent arm, the beam index is embedded into the context, enabling BKC-UCB to exploit correlations among beams to accelerate convergence. Furthermore, an event-triggered information sharing mechanism is incorporated into BKC-UCB, enabling information exchange only when significant explorations are conducted to improve learning efficiency with limited communication overhead.
【21】Structured Latent Dynamics in Wireless CSI via Homomorphic World Models
标题：基于同态世界模型的无线CSI结构化潜在动态
链接：https://arxiv.org/abs/2603.20048
备注：ACCEPTED FOR PUBLICATION IN IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS (ICC) 2026
摘要：We introduce a self-supervised framework for learning predictive and structured representations of wireless channels by modeling the temporal evolution of channel state information (CSI) in a compact latent space. Our method casts the problem as a world modeling task and leverages the Joint Embedding Predictive Architecture (JEPA) to learn action-conditioned latent dynamics from CSI trajectories. To promote geometric consistency and compositionality, we parameterize transitions using homomorphic updates derived from Lie algebra, yielding a structured latent space that reflects spatial layout and user motion. Evaluations on the DICHASUS dataset show that our approach outperforms strong baselines in preserving topology and forecasting future embeddings across unseen environments. The resulting latent space enables metrically faithful channel charts, offering a scalable foundation for downstream applications such as mobility-aware scheduling, localization, and wireless scene understanding.
【22】A two-step sequential approach for hyperparameter selection in finite context models
标题：有限上下文模型中超参数选择的两步顺序方法
链接：https://arxiv.org/abs/2603.19736
摘要：Finite-context models (FCMs) are widely used for compressing symbolic sequences such as DNA, where predictive performance depends critically on the context length k and smoothing parameter α. In practice, these hyperparameters are typically selected through exhaustive search, which is computationally expensive and scales poorly with model complexity. This paper proposes a statistically grounded two-step sequential approach for efficient hyperparameter selection in FCMs. The key idea is to decompose the joint optimization problem into two independent stages. First, the context length k is estimated using categorical serial dependence measures, including Cramér's ν, Cohen's \k{appa} and partial mutual information (pami). Second, the smoothing parameter α is estimated via maximum likelihood conditional on the selected context length k. Simulation experiments were conducted on synthetic symbolic sequences generated by FCMs across multiple (k, α) configurations, considering a four-letter alphabet and different sample sizes. Results show that the dependence measures are substantially more sensitive to variations in k than in α, supporting the sequential estimation strategy. As expected, the accuracy of the hyperparameter estimation improves with increasing sample size. Furthermore, the proposed method achieves compression performance comparable to exhaustive grid search in terms of average bitrate (bits per symbol), while substantially reducing computational cost. Overall, the results on simulated data show that the proposed sequential approach is a practical and computationally efficient alternative to exhaustive hyperparameter tuning in FCMs.
【23】Learning to Bet for Horizon-Aware Anytime-Valid Testing
标题：学习投注地平线感知随时有效的测试
链接：https://arxiv.org/abs/2603.19551
备注：21 pages, 27 figures
摘要：We develop horizon-aware anytime-valid tests and confidence sequences for bounded means under a strict deadline $N$. Using the betting/e-process framework, we cast horizon-aware betting as a finite-horizon optimal control problem with state space $(t, \log W_t)$, where $t$ is the time and $W_t$ is the test martingale value. We first show that in certain interior regions of the state space, policies that deviate significantly from Kelly betting are provably suboptimal, while Kelly betting reaches the threshold with high probability. We then identify sufficient conditions showing that outside this region, more aggressive betting than Kelly can be better if the bettor is behind schedule, and less aggressive can be better if the bettor is ahead. Taken together these results suggest a simple phase diagram in the $(t, \log W_t)$ plane, delineating regions where Kelly, fractional Kelly, and aggressive betting may be preferable. Guided by this phase diagram, we introduce a Deep Reinforcement Learning approach based on a universal Deep Q-Network (DQN) agent that learns a single policy from synthetic experience and maps simple statistics of past observations to bets across horizons and null values. In limited-horizon experiments, the learned DQN policy yields state-of-the-art results.
【24】Near-Equivalent Q-learning Policies for Dynamic Treatment Regimes
标题：动态治疗方案的接近等效Q学习策略
链接：https://arxiv.org/abs/2603.19440
备注：13 pages, 2 figures
摘要：Precision medicine aims to tailor therapeutic decisions to individual patient characteristics. This objective is commonly formalized through dynamic treatment regimes, which use statistical and machine learning methods to derive sequential decision rules adapted to evolving clinical information. In most existing formulations, these approaches produce a single optimal treatment at each stage, leading to a unique decision sequence. However, in many clinical settings, several treatment options may yield similar expected outcomes, and focusing on a single optimal policy may conceal meaningful alternatives. We extend the Q-learning framework for retrospective data by introducing a worst-value tolerance criterion controlled by a hyperparameter $\varepsilon$, which specifies the maximum acceptable deviation from the optimal expected value. Rather than identifying a single optimal policy, the proposed approach constructs sets of $\varepsilon$-optimal policies whose performance remains within a controlled neighborhood of the optimum. This formulation shifts Q-learning from a vector-valued representation to a matrix-valued one, allowing multiple admissible value functions to coexist during backward recursion. The approach yields families of near-equivalent treatment strategies and explicitly identifies regions of treatment indifference where several decisions achieve comparable outcomes. We illustrate the framework in two settings: a single-stage problem highlighting indifference regions around the decision boundary, and a multi-stage decision process based on a simulated oncology model describing tumor size and treatment toxicity dynamics.
【25】Joint Return and Risk Modeling with Deep Neural Networks for Portfolio Construction
标题：利用深度神经网络进行联合回报和风险建模用于投资组合构建
链接：https://arxiv.org/abs/2603.19288
摘要：Portfolio construction traditionally relies on separately estimating expected returns and covariance matrices using historical statistics, often leading to suboptimal allocation under time-varying market conditions. This paper proposes a joint return and risk modeling framework based on deep neural networks that enables end-to-end learning of dynamic expected returns and risk structures from sequential financial data. Using daily data from ten large-cap US equities spanning 2010 to 2024, the proposed model is evaluated across return prediction, risk estimation, and portfolio-level performance. Out-of-sample results during 2020 to 2024 show that the deep forecasting model achieves competitive predictive accuracy (RMSE = 0.0264) with economically meaningful directional accuracy (51.9%). More importantly, the learned representation effectively captures volatility clustering and regime shifts. When integrated into portfolio optimization, the proposed Neural Portfolio strategy achieves an annual return of 36.4% and a Sharpe ratio of 0.91, outperforming equal weight and historical mean-variance benchmarks in terms of risk-adjusted performance. These findings demonstrate that jointly modeling return and covariance dynamics can provide consistent improvements over traditional allocation approaches. The framework offers a scalable and practical alternative for data-driven portfolio construction under nonstationary market conditions.
其他(29篇)
【1】From Masks to Pixels and Meaning: A New Taxonomy, Benchmark, and Metrics for VLM Image Tampering
标题：从面具到像素和意义：VLM图像篡改的新分类、基准和收件箱
链接：https://arxiv.org/abs/2603.20193
备注：Code and data at: https://github.com/VILA-Lab/PIXAR (Accepted in CVPR 2026 Findings, but not opted in)
摘要：Existing tampering detection benchmarks largely rely on object masks, which severely misalign with the true edit signal: many pixels inside a mask are untouched or only trivially modified, while subtle yet consequential edits outside the mask are treated as natural. We reformulate VLM image tampering from coarse region labels to a pixel-grounded, meaning and language-aware task. First, we introduce a taxonomy spanning edit primitives (replace/remove/splice/inpaint/attribute/colorization, etc.) and their semantic class of tampered object, linking low-level changes to high-level understanding. Second, we release a new benchmark with per-pixel tamper maps and paired category supervision to evaluate detection and classification within a unified protocol. Third, we propose a training framework and evaluation metrics that quantify pixel-level correctness with localization to assess confidence or prediction on true edit intensity, and further measure tamper meaning understanding via semantics-aware classification and natural language descriptions for the predicted regions. We also re-evaluate the existing strong segmentation/localization baselines on recent strong tamper detectors and reveal substantial over- and under-scoring using mask-only metrics, and expose failure modes on micro-edits and off-mask changes. Our framework advances the field from masks to pixels, meanings and language descriptions, establishing a rigorous standard for tamper localization, semantic classification and description. Code and benchmark data are available at https://github.com/VILA-Lab/PIXAR.
【2】MeanFlow Meets Control: Scaling Sampled-Data Control for Swarms
标题：MeanFlow满足控制：扩展群的采样数据控制
链接：https://arxiv.org/abs/2603.20189
摘要：Steering large-scale swarms in only a few control updates is challenging because real systems operate in sampled-data form: control inputs are updated intermittently and applied over finite intervals. In this regime, the natural object is not an instantaneous velocity field, but a finite-window control quantity that captures the system response over each sampling interval. Inspired by MeanFlow, we introduce a control-space learning framework for swarm steering under linear time-invariant dynamics. The learned object is the coefficient that parameterizes the finite-horizon minimum-energy control over each interval. We show that this coefficient admits both an integral representation and a local differential identity along bridge trajectories, which leads to a simple stop-gradient training objective. At implementation time, the learned coefficient is used directly in sampled-data updates, so the prescribed dynamics and actuation map are respected by construction. The resulting framework provides a scalable approach to few-step swarm steering that is consistent with the sampled-data structure of real control systems.
【3】Revisiting Gene Ontology Knowledge Discovery with Hierarchical Feature Selection and Virtual Study Group of AI Agents
标题：利用分层特征选择和人工智能代理虚拟研究小组重新审视基因本体知识发现
链接：https://arxiv.org/abs/2603.20132
摘要：Large language models have achieved great success in multiple challenging tasks, and their capacity can be further boosted by the emerging agentic AI techniques. This new computing paradigm has already started revolutionising the traditional scientific discovery pipelines. In this work, we propose a novel agentic AI-based knowledge discovery-oriented virtual study group that aims to extract meaningful ageing-related biological knowledge considering highly ageing-related Gene Ontology terms that are selected by hierarchical feature selection methods. We investigate the performance of the proposed agentic AI framework by considering four different model organisms' ageing-related Gene Ontology terms and validate the biological findings by reviewing existing research articles. It is found that the majority of the AI agent-generated scientific claims can be supported by existing literatures and the proposed internal mechanisms of the virtual study group also play an important role in the designed agentic AI-based knowledge discovery framework.
【4】GO-GenZip: Goal-Oriented Generative Sampling and Hybrid Compression
标题：GO-GenZip：面向目标的生成采样和混合压缩
链接：https://arxiv.org/abs/2603.20109
摘要：Current network data telemetry pipelines consist of massive streams of fine-grained Key Performance Indicators (KPIs) from multiple distributed sources towards central aggregators, making data storage, transmission, and real-time analysis increasingly unsustainable. This work presents a generative AI (GenAI)-driven sampling and hybrid compression framework that redesigns network telemetry from a goal-oriented perspective. Unlike conventional approaches that passively compress fully observed data, our approach jointly optimizes what to observe and how to encode it, guided by the relevance of information to downstream tasks. The framework integrates adaptive sampling policies, using adaptive masking techniques, with generative modeling to identify patterns and preserve critical features across temporal and spatial dimensions. The selectively acquired data are further processed through a hybrid compression scheme that combines traditional lossless coding with GenAI-driven, lossy compression. Experimental results on real network datasets demonstrate over 50$\%$ reductions in sampling and data transfer costs, while maintaining comparable reconstruction accuracy and goal-oriented analytical fidelity in downstream tasks.
【5】Federated Hyperdimensional Computing for Resource-Constrained Industrial IoT
标题：面向资源受限工业物联网的联邦超维计算
链接：https://arxiv.org/abs/2603.20037
备注：Submitted to the IEEE for possible publication
摘要：In the Industrial Internet of Things (IIoT) systems, edge devices often operate under strict constraints in memory, compute capability, and wireless bandwidth. These limitations challenge the deployment of advanced data analytics tasks, such as predictive and prescriptive maintenance. In this work, we explore hyperdimensional computing (HDC) as a lightweight learning paradigm for resource-constrained IIoT. Conventional centralized HDC leverages the properties of high-dimensional vector spaces to enable energy-efficient training and inference. We integrate this paradigm into a federated learning (FL) framework where devices exchange only prototype representations, which significantly reduces communication overhead. Our numerical results highlight the potential of federated HDC to support collaborative learning in IIoT with fast convergence speed and communication efficiency. These results indicate that HDC represents a lightweight and resilient framework for distributed intelligence in large-scale and resource-constrained IIoT environments.
【6】AgenticRS-EnsNAS: Ensemble-Decoupled Self-Evolving Architecture Search
标题：统计学RS-EnsNAS：集成脱钩的自进化架构搜索
链接：https://arxiv.org/abs/2603.20014
摘要：Neural Architecture Search (NAS) deployment in industrial production systems faces a fundamental validation bottleneck: verifying a single candidate architecture pi requires evaluating the deployed ensemble of M models, incurring prohibitive O(M) computational cost per candidate. This cost barrier severely limits architecture iteration frequency in real-world applications where ensembles (M=50-200) are standard for robustness. This work introduces Ensemble-Decoupled Architecture Search, a framework that leverages ensemble theory to predict system-level performance from single-learner evaluation. We establish the Ensemble-Decoupled Theory with a sufficient condition for monotonic ensemble improvement under homogeneity assumptions: a candidate architecture pi yields lower ensemble error than the current baseline if rho(pi) < rho(pi_old) - (M / (M - 1)) * (Delta E(pi) / sigma^2(pi)), where Delta E, rho, and sigma^2 are estimable from lightweight dual-learner training. This decouples architecture search from full ensemble training, reducing per-candidate search cost from O(M) to O(1) while maintaining O(M) deployment cost only for validated winners. We unify solution strategies across pipeline continuity: (1) closed-form optimization for tractable continuous pi (exemplified by feature bagging in CTR prediction), (2) constrained differentiable optimization for intractable continuous pi, and (3) LLM-driven search with iterative monotonic acceptance for discrete pi. The framework reveals two orthogonal improvement mechanisms -- base diversity gain and accuracy gain -- providing actionable design principles for industrial-scale NAS. All theoretical derivations are rigorous with detailed proofs deferred to the appendix. Comprehensive empirical validation will be included in the journal extension of this work.
【7】A Super Fast K-means for Indexing Vector Embeddings
标题：一种用于向量嵌入索引的超快速K-means算法
链接：https://arxiv.org/abs/2603.20009
摘要：We present SuperKMeans: a k-means variant designed for clustering collections of high-dimensional vector embeddings. SuperKMeans' clustering is up to 7x faster than FAISS and Scikit-Learn on modern CPUs and up to 4x faster than cuVS on GPUs (Figure 1), while maintaining the quality of the resulting centroids for vector similarity search tasks. SuperKMeans acceleration comes from reducing data-access and compute overhead by reliably and efficiently pruning dimensions that are not needed to assign a vector to a centroid. Furthermore, we present Early Termination by Recall, a novel mechanism that early-terminates k-means when the quality of the centroids for retrieval tasks stops improving across iterations. In practice, this further reduces runtimes without compromising retrieval quality. We open-source our implementation at https://github.com/cwida/SuperKMeans
【8】TAPAS: Efficient Two-Server Asymmetric Private Aggregation Beyond Prio(+)
标题：TAPAS：超越Prio（+）的高效两服务器非对称私有聚合
链接：https://arxiv.org/abs/2603.19949
摘要 ：Privacy-preserving aggregation is a cornerstone for AI systems that learn from distributed data without exposing individual records, especially in federated learning and telemetry. Existing two-server protocols (e.g., Prio and successors) set a practical baseline by validating inputs while preventing any single party from learning users' values, but they impose symmetric costs on both servers and communication that scales with the per-client input dimension $L$. Modern learning tasks routinely involve dimensionalities $L$ in the tens to hundreds of millions of model parameters. We present TAPAS, a two-server asymmetric private aggregation scheme that addresses these limitations along four dimensions: (i) no trusted setup or preprocessing, (ii) server-side communication that is independent of $L$ (iii) post-quantum security based solely on standard lattice assumptions (LWE, SIS), and (iv) stronger robustness with identifiable abort and full malicious security for the servers. A key design choice is intentional asymmetry: one server bears the $O(L)$ aggregation and verification work, while the other operates as a lightweight facilitator with computation independent of $L$. This reduces total cost, enables the secondary server to run on commodity hardware, and strengthens the non-collusion assumption of the servers. One of our main contributions is a suite of new and efficient lattice-based zero-knowledge proofs; to our knowledge, we are the first to establish privacy and correctness with identifiable abort in the two-server setting.
【9】Discovery of Decision Synchronization Patterns from Event Logs
标题：从事件查找器发现决策同步模式
链接：https://arxiv.org/abs/2603.19879
摘要：Synchronizing decisions between running cases in business processes facilitates fair and efficient use of resources, helps prioritize the most valuable cases, and prevents unnecessary waiting. Consequently, decision synchronization patterns are regularly built into processes, in the form of mechanisms that temporarily delay one case to favor another. These decision mechanisms therefore consider properties of multiple cases at once, rather than just the properties of a single case; an aspect that is rarely addressed by current process discovery techniques. To address this gap, this paper proposes an approach for discovering decision synchronization patterns inspired by supply chain processes. These decision synchronization patterns take the form of specific process constructs combined with a constraint that determines which particular case to execute. We describe, formalize and demonstrate how the constraint for four such patterns can be discovered. We evaluate our approach in two artificial scenarios. First, with four separate process models each containing a single decision synchronization pattern, i.e., we demonstrate that our approach can discover every type of pattern when only this one type is present. Second, we consider a process model containing all four decision synchronization patterns to show generalizability of the approach to more complex problems. For both scenarios, we could reliably retrieve the expected patterns.
【10】On the Dynamics & Transferability of Latent Generalization during Memorization
标题：论企业化过程中潜在概括的动态性和可转移性
链接：https://arxiv.org/abs/2603.19865
摘要：Deep networks have been known to have extraordinary generalization abilities, via mechanisms that aren't yet well understood. It is also known that upon shuffling labels in the training data to varying degrees, deep networks, trained with standard methods, can still achieve perfect or high accuracy on this corrupted training data. This phenomenon is called memorization, and typically comes at the cost of poorer generalization to true labels. Our recent work has demonstrated, that the internal representations of such models retain significantly better latent generalization abilities than is directly apparent from the model. In particular, it has been shown that such latent generalization can be recovered via simple probes (called MASC probes) on the layer-wise representations of the model. However, the origin and dynamics over training of this latent generalization during memorization is not well understood. Here, we track the training dynamics, empirically, and find that latent generalization abilities largely peak early in training, with model generalization. Next, we investigate to what extent the specific nature of the MASC probe is critical for our ability to extract latent generalization from the model's layerwise outputs. To this end, we first examine the mathematical structure of the MASC probe and show that it is a quadratic classifier, i.e. is non-linear. This brings up the question of the extent to which this latent generalization might be linearly decodable from layerwise outputs. To investigate this, we designed a new linear probe for this setting. Next, we consider the question of whether it is possible to transfer latent generalization to model generalization by directly editing model weights. To this end, we devise a way to transfer the latent generalization present in last-layer representations to the model using the new linear probe.
【11】IsoCLIP: Decomposing CLIP Projectors for Efficient Intra-modal Alignment
标题：IsoCLIP：分解CLIP投影仪以实现高效的模式内对齐
链接：https://arxiv.org/abs/2603.19862
备注：Accepted at CVPR2026
摘要：Vision-Language Models like CLIP are extensively used for inter-modal tasks which involve both visual and text modalities. However, when the individual modality encoders are applied to inherently intra-modal tasks like image-to-image retrieval, their performance suffers from the intra-modal misalignment. In this paper we study intra-modal misalignment in CLIP with a focus on the role of the projectors that map pre-projection image and text embeddings into the shared embedding space. By analyzing the form of the cosine similarity applied to projected features, and its interaction with the contrastive CLIP loss, we show that there is an inter-modal operator responsible for aligning the two modalities during training, and a second, intra-modal operator that only enforces intra-modal normalization but does nothing to promote intra-modal alignment. Via spectral analysis of the inter-modal operator, we identify an approximately isotropic subspace in which the two modalities are well-aligned, as well as anisotropic directions specific to each modality. We demonstrate that this aligned subspace can be directly obtained from the projector weights and that removing the anisotropic directions improves intra-modal alignment. Our experiments on intra-modal retrieval and classification benchmarks show that our training-free method reduces intra-modal misalignment, greatly lowers latency, and outperforms existing approaches across multiple pre-trained CLIP-like models. The code is publicly available at: https://github.com/simomagi/IsoCLIP.
【12】Scale-Dependent Radial Geometry and Metric Mismatch in Wasserstein Propagation for Reverse Diffusion
标题
：反向扩散Wasserstein传播中的尺度相关的辐射几何和度量不匹配
链接：https://arxiv.org/abs/2603.19670
摘要：Existing analyses of reverse diffusion often propagate sampling error in the Euclidean geometry underlying \(\Wtwo\) along the entire reverse trajectory. Under weak log-concavity, however, Gaussian smoothing can create contraction first at large separations while short separations remain non-dissipative. The first usable contraction is therefore radial rather than Euclidean, creating a metric mismatch between the geometry that contracts early and the geometry in which the terminal error is measured. We formalize this mismatch through an explicit radial lower profile for the learned reverse drift. Its far-field limit gives a contraction reserve, its near-field limit gives the Euclidean load governing direct \(\Wtwo\) propagation, and admissible switch times are characterized by positivity of the reserve on the remaining smoothing window. We exploit this structure with a one-switch routing argument. Before the switch, reflection coupling yields contraction in a concave transport metric adapted to the radial profile. At the switch, we convert once from this metric back to \(\Wtwo\) under a \(p\)-moment budget, and then propagate the converted discrepancy over the remaining short window in Euclidean geometry. For discretizations of the learned reverse SDE under \(L^2\) score-error control, a one-sided Lipschitz condition of score error, and standard well-posedness and coupling hypotheses, we obtain explicit non-asymptotic end-to-end \(\Wtwo\) guarantees, a scalar switch-selection objective, and a sharp structural limit on the conversion exponent within the affine-tail concave class.
【13】Alternating Diffusion for Proximal Sampling with Zeroth Order Queries
标题：具有零级搜索器的近端采样的交替扩散
链接：https://arxiv.org/abs/2603.19633
备注：Accepted to ICLR2026
摘要：This work introduces a new approximate proximal sampler that operates solely with zeroth-order information of the potential function. Prior theoretical analyses have revealed that proximal sampling corresponds to alternating forward and backward iterations of the heat flow. The backward step was originally implemented by rejection sampling, whereas we directly simulate the dynamics. Unlike diffusion-based sampling methods that estimate scores via learned models or by invoking auxiliary samplers, our method treats the intermediate particle distribution as a Gaussian mixture, thereby yielding a Monte Carlo score estimator from directly samplable distributions. Theoretically, when the score estimation error is sufficiently controlled, our method inherits the exponential convergence of proximal sampling under isoperimetric conditions on the target distribution. In practice, the algorithm avoids rejection sampling, permits flexible step sizes, and runs with a deterministic runtime budget. Numerical experiments demonstrate that our approach converges rapidly to the target distribution, driven by interactions among multiple particles and by exploiting parallel computation.
【14】K-GMRF: Kinetic Gauss-Markov Random Field for First-Principles Covariance Tracking on Lie Groups
标题：K-GMRF：李群第一原理协方差跟踪的动态高斯-马尔科夫随机场
链接：https://arxiv.org/abs/2603.19601
备注：33 pages, 13 figures
摘要：Tracking non-stationary covariance matrices is fundamental to vision yet hindered by existing estimators that either neglect manifold constraints or rely on first-order updates, incurring inevitable phase lag during rapid evolution. We propose K-GMRF, an online, training-free framework for covariance tracking that reformulates the problem as forced rigid-body motion on Lie groups. Derived from the Euler-Poincaré equations, our method interprets observations as torques driving a latent angular velocity, propagated via a structure-preserving symplectic integrator. We theoretically prove that this second-order dynamics achieves zero steady-state error under constant rotation, strictly superior to the proportional lag of first-order baselines. Validation across three domains demonstrates robust tracking fidelity: (i) on synthetic ellipses, K-GMRF reduces angular error by 30x compared to Riemannian EMA while maintaining stability at high speeds; (ii) on SO(3) stabilization with 20% dropout, it decreases geodesic error from 29.4° to 9.9°; and (iii) on OTB motion-blur sequences, it improves loU from 0.55 to 0.74 on BlurCar2 with a 96% success rate. As a fully differentiable symplectic module, K-GMRF provides a plug-and-play geometric prior for data-constrained scenarios and an interpretable layer within modern deep architectures.
【15】SurfaceXR: Fusing Smartwatch IMUs and Egocentric Hand Pose for Seamless Surface Interactions
标题：SurfaceXR：融合Smartwatch IMU和以自我为中心的手势，实现无缝的表面交互
链接：https://arxiv.org/abs/2603.19529
备注：Accepted to IEEE VR 2026 as a TVCG journal paper
摘要：Mid-air gestures in Extended Reality (XR) often cause fatigue and imprecision. Surface-based interactions offer improved accuracy and comfort, but current egocentric vision methods struggle due to hand tracking challenges and unreliable surface plane estimation. We introduce SurfaceXR, a sensor fusion approach combining headset-based hand tracking with smartwatch IMU data to enable robust inputs on everyday surfaces. Our insight is that these modalities are complementary: hand tracking provides 3D positional data while IMUs capture high-frequency motion. A 21-participant study validates SurfaceXR's effectiveness for touch tracking and 8-class gesture recognition, demonstrating significant improvements over single-modality approaches.
【16】Teaching an Agent to Sketch One Part at a Time
标题：教代理一次绘制一个部分
链接：https://arxiv.org/abs/2603.19500
摘要：We develop a method for producing vector sketches one part at a time. To do this, we train a multi-modal language model-based agent using a novel multi-turn process-reward reinforcement learning following supervised fine-tuning. Our approach is enabled by a new dataset we call ControlSketch-Part, containing rich part-level annotations for sketches, obtained using a novel, generic automatic annotation pipeline that segments vector sketches into semantic parts and assigns paths to parts with a structured multi-stage labeling process. Our results indicate that incorporating structured part-level data and providing agent with the visual feedback through the process enables interpretable, controllable, and locally editable text-to-vector sketch generation.
【17】TRACE: Trajectory Recovery with State Propagation Diffusion for Urban Mobility
标题：TRACE：城市流动性国家传播扩散的轨迹恢复
链接：https://arxiv.org/abs/2603.19474
备注：This article is accepted by WWW 2026, Dubai, United Arab Emirates
摘要：High-quality GPS trajectories are essential for location-based web services and smart city applications, including navigation, ride-sharing and delivery. However, due to low sampling rates and limited infrastructure coverage during data collection, real-world trajectories are often sparse and feature unevenly distributed location points. Recovering these trajectories into dense and continuous forms is essential but challenging, given their complex and irregular spatio-temporal patterns. In this paper, we introduce a novel diffusion model for trajectory recovery named TRACE, which reconstruct dense and continuous trajectories from sparse and incomplete inputs. At the core of TRACE, we propose a State Propagation Diffusion Model (SPDM), which integrates a novel memory mechanism, so that during the denoising process, TRACE can retain and leverage intermediate results from previous steps to effectively reconstruct those hard-to-recover trajectory segments. Extensive experiments on multiple real-world datasets show that TRACE outperforms the state-of-the-art, offering $>$26\% accuracy improvement without significant inference overhead. Our work strengthens the foundation for mobile and web-connected location services, advancing the quality and fairness of data-driven urban applications. Code is available at: https://github.com/JinmingWang/TRACE
【18】Scalable Prompt Routing via Fine-Grained Latent Task Discovery
标题：通过细粒度潜在任务发现的可扩展提示路由
链接：https://arxiv.org/abs/2603.19415
摘要：Prompt routing dynamically selects the most appropriate large language model from a pool of candidates for each query, optimizing performance while managing costs. As model pools scale to include dozens of frontier models with narrow performance gaps, existing approaches face significant challenges: manually defined task taxonomies cannot capture fine-grained capability distinctions, while monolithic routers struggle to differentiate subtle differences across diverse tasks. We propose a two-stage routing architecture that addresses these limitations through automated fine-grained task discovery and task-aware quality estimation. Our first stage employs graph-based clustering to discover latent task types and trains a classifier to assign prompts to discovered tasks. The second stage uses a mixture-of-experts architecture with task-specific prediction heads for specialized quality estimates. At inference, we aggregate predictions from both stages to balance task-level stability with prompt-specific adaptability. Evaluated on 10 benchmarks with 11 frontier models, our method consistently outperforms existing baselines and surpasses the strongest individual model while incurring less than half its cost.
【19】Beyond Weighted Summation: Learnable Nonlinear Aggregation Functions for Robust Artificial Neurons
标题：超越加权总和：鲁棒人工神经元的可学习非线性聚集函数
链接：https://arxiv.org/abs/2603.19344
备注：7 pages, 2 tables
摘要：Weighted summation has remained the default input aggregation mechanism in artificial neurons since the earliest neural network models. While computationally efficient, this design implicitly behaves like a mean-based estimator and is therefore sensitive to noisy or extreme inputs. This paper investigates whether replacing fixed linear aggregation with learnable nonlinear alternatives can improve neural network robustness without sacrificing trainability. Two differentiable aggregation mechanisms are introduced: an F-Mean neuron based on a learnable power-weighted aggregation rule, and a Gaussian Support neuron based on distance-aware affinity weighting. To preserve the optimisation stability of standard neurons, hybrid neurons are proposed that interpolate between linear and nonlinear aggregation through a learnable blending parameter. Evaluated in multilayer perceptrons and convolutional neural networks on CIFAR-10 and a noisy CIFAR-10 variant with additive Gaussian corruption, hybrid neurons consistently improve robustness under noise while F-Mean hybrids also yield modest gains on clean data. The three-way hybrid achieves robustness scores of up to 0.991 compared to 0.890 for the standard baseline, and learned parameters converge consistently to sub-linear aggregation (p $\approx$ 0.43--0.50) and high novelty utilisation ($α$ $\approx$ 0.69--0.79). These findings suggest that neuron-level aggregation is a meaningful and underexplored design dimension for building more noise-tolerant neural networks.
【20】Ternary Gamma Semirings: From Neural Implementation to Categorical Foundations
标题：三元伽玛半林：从神经实现到分类基础
链接：https://arxiv.org/abs/2603.19317
摘要 ：This paper establishes a theoretical framework connecting neural network learning with abstract algebraic structures. We first present a minimal counterexample demonstrating that standard neural networks completely fail on compositional generalization tasks (0% accuracy). By introducing a logical constraint -- the Ternary Gamma Semiring -- the same architecture learns a perfectly structured feature space, achieving 100% accuracy on novel combinations. We prove that this learned feature space constitutes a finite commutative ternary $Γ$-semiring, whose ternary operation implements the majority vote rule. Comparing with the recently established classification of Gokavarapu et al., we show that this structure corresponds precisely to the Boolean-type ternary $Γ$-semiring with $|T|=4$, $|Γ|=1$}, which is unique up to isomorphism in their enumeration. Our findings reveal three profound conclusions: (i) the success of neural networks can be understood as an approximation of mathematically ``natural'' structures; (ii) learned representations generalize because they internalize algebraic axioms (symmetry, idempotence, majority property); (iii) logical constraints guide networks to converge to these canonical forms. This work provides a rigorous mathematical framework for understanding neural network generalization and inaugurates the new interdisciplinary direction of Computational $Γ$-Algebra.
【21】GT-Space: Enhancing Heterogeneous Collaborative Perception with Ground Truth Feature Space
标题：GT-Space：利用地面真相特征空间增强异类协作感知
链接：https://arxiv.org/abs/2603.19308
摘要：In autonomous driving, multi-agent collaborative perception enhances sensing capabilities by enabling agents to share perceptual data. A key challenge lies in handling {\em heterogeneous} features from agents equipped with different sensing modalities or model architectures, which complicates data fusion. Existing approaches often require retraining encoders or designing interpreter modules for pairwise feature alignment, but these solutions are not scalable in practice. To address this, we propose {\em GT-Space}, a flexible and scalable collaborative perception framework for heterogeneous agents. GT-Space constructs a common feature space from ground-truth labels, providing a unified reference for feature alignment. With this shared space, agents only need a single adapter module to project their features, eliminating the need for pairwise interactions with other agents. Furthermore, we design a fusion network trained with contrastive losses across diverse modality combinations. Extensive experiments on simulation datasets (OPV2V and V2XSet) and a real-world dataset (RCooper) demonstrate that GT-Space consistently outperforms baselines in detection accuracy while delivering robust performance. Our code will be released at https://github.com/KingScar/GT-Space.
【22】Autonoma: A Hierarchical Multi-Agent Framework for End-to-End Workflow Automation
标题：Autonoma：用于端到端工作流程自动化的分层多代理框架
链接：https://arxiv.org/abs/2603.19270
备注：26 Pages, 3 Figures
摘要：The increasing complexity of user demands necessitates automation frameworks that can reliably translate open-ended instructions into robust, multi-step workflows. Current monolithic agent architectures often struggle with the challenges of scalability, error propagation, and maintaining focus across diverse tasks. This paper introduces Autonoma, a structured, hierarchical multi-agent framework designed for end-to-end workflow automation from natural language prompts. Autonoma employs a principled, multi-tiered architecture where a high-level Coordinator validates user intent, a Planner generates structured workflows, and a Supervisor dynamically manages the execution by orchestrating a suite of modular, specialized agents (e.g., for web browsing, coding, file management). This clear separation between orchestration logic and specialized execution ensures robustness through active monitoring and error handling, while enabling extensibility by allowing new capabilities to be integrated as plug-and-play agents without modifying the core engine. Implemented as a fully functional system operating within a secure LAN environment, Autonoma addresses critical data privacy and reliability concerns. The system is further engineered for inclusivity, accepting multi-modal input (text, voice, image, files) and supporting both English and Arabic. Autonoma achieved a 97% task completion rate and a 98% successful agent handoff rate, confirming its operational reliability and efficient collaboration.
【23】MAPLE: Metadata Augmented Private Language Evolution
标题：MAPLE：元数据增强私人语言进化
链接：https://arxiv.org/abs/2603.19258
备注：Preliminary work
摘要：While differentially private (DP) fine-tuning of large language models (LLMs) is a powerful tool, it is often computationally prohibitive or infeasible when state-of-the-art models are only accessible via proprietary APIs. In such settings, generating DP synthetic data has emerged as a crucial alternative, offering the added benefits of arbitrary reuse across downstream tasks and transparent exploratory data analysis without the opaque constraints of a model's parameter space. Private Evolution (PE) is a promising API-based framework for this goal; however, its performance critically depends on initialization. When the private data distribution deviates substantially from the foundation model's pre-training priors--particularly in highly specialized domains--PE frequently struggles to align with the target data, resulting in degraded utility, poor convergence, and inefficient API usage. To address this initialization bottleneck, we propose Metadata Augmented Private Language Evolution (MAPLE). MAPLE leverages differentially private tabular metadata extraction and in-context learning to effectively ground the initial synthetic distribution in the target domain. Extensive experiments on challenging, domain-specific text generation tasks demonstrate that MAPLE achieves a significantly more favorable privacy-utility trade-off, converges faster, and drastically reduces API costs compared to previous PE methods.
【24】The IJCNN 2025 Review Process
标题：IJCNN 2025审查流程
链接：https://arxiv.org/abs/2603.19244
摘要：The International Joint Conference on Neural Networks (IJCNN) is the premier international conference in the area of neural networks theory, analysis, and applications. The 2025 edition of the conference comprised 5,526 paper submissions, 7,877 active reviewers, 426 area chairs, 2,152 accepted papers, and more than 2,300 attendees. This represents a growth of about 100% in terms of submissions, 200% in terms of reviewers, and over 50% in terms of attendees as compared to the previous edition. In this paper, we describe several key aspects of the whole review process, including a strategy for ranking the scores provided by the reviewers by evaluating a score index and a calibrated version used experimentally to remove reviewer-specific bias from reviews.
【25】Survey of Various Fuzzy and Uncertain Decision-Making Methods
标题：各种模糊不确定决策方法综述
链接：https://arxiv.org/abs/2603.15709
备注：Book. Publisher: Neutrosophic Science International Association (NSIA) Publishing House. ISBN: 978-1-59973-883-3. 446 pages
摘要：Decision-making in real applications is often affected by vagueness, incomplete information, heterogeneous data, and conflicting expert opinions. This survey reviews uncertainty-aware multi-criteria decision-making (MCDM) and organizes the field into a concise, task-oriented taxonomy. We summarize problem-level settings (discrete, group/consensus, dynamic, multi-stage, multi-level, multiagent, and multi-scenario), weight elicitation (subjective and objective schemes under fuzzy/linguistic inputs), and inter-criteria structure and causality modelling. For solution procedures, we contrast compensatory scoring methods, distance-to-reference and compromise approaches, and non-compensatory outranking frameworks for ranking or sorting. We also outline rule/evidence-based and sequential decision models that produce interpretable rules or policies. The survey highlights typical inputs, core computational steps, and primary outputs, and provides guidance on choosing methods according to robustness, interpretability, and data availability. It concludes with open directions on explainable uncertainty integration, stability, and scalability in large-scale and dynamic decision environments.
【26】AI Agents Can Already Autonomously Perform Experimental High Energy Physics
标题：人工智能代理已经可以自主进行实验性高能物理
链接：https://arxiv.org/abs/2603.20179
摘要：Large language model-based AI agents are now able to autonomously execute substantial portions of a high energy physics (HEP) analysis pipeline with minimal expert-curated input. Given access to a HEP dataset, an execution framework, and a corpus of prior experimental literature, we find that Claude Code succeeds in automating all stages of a typical analysis: event selection, background estimation, uncertainty quantification, statistical inference, and paper drafting. We argue that the experimental HEP community is underestimating the current capabilities of these systems, and that most proposed agentic workflows are too narrowly scoped or scaffolded to specific analysis structures. We present a proof-of-concept framework, Just Furnish Context (JFC), that integrates autonomous analysis agents with literature-based knowledge retrieval and multi-agent review, and show that this is sufficient to plan, execute, and document a credible high energy physics analysis. We demonstrate this by conducting analyses on open data from ALEPH, DELPHI, and CMS to perform electroweak, QCD, and Higgs boson measurements. Rather than replacing physicists, these tools promise to offload the repetitive technical burden of analysis code development, freeing researchers to focus on physics insight, truly novel method development, and rigorous validation. Given these developments, we advocate for new strategies for how the community trains students, organizes analysis efforts, and allocates human expertise.
【27】Structural Controllability of Large-Scale Hypergraphs
标题：大规模超图的结构可控性
链接：https://arxiv.org/abs/2603.19955
备注：14 pages, 4 figures, 1 table
摘要：Controlling real-world networked systems, including ecological, biomedical, and engineered networks that exhibit higher-order interactions, remains challenging due to inherent nonlinearities and large system scales. Despite extensive studies on graph controllability, the controllability properties of hypergraphs remain largely underdeveloped. Existing results focus primarily on exact controllability, which is often impractical for large-scale hypergraphs. In this article, we develop a structural controllability framework for hypergraphs by modeling hypergraph dynamics as polynomial dynamical systems. In particular, we extend classical notions of accessibility and dilation from linear graph-based systems to polynomial hypergraph dynamics and establish a hypergraph-based criterion under which the topology guarantees satisfaction of classical Lie-algebraic and Kalman-type rank conditions for almost all parameter choices. We further derive a topology-based lower bound on the minimum number of driver nodes required for structural controllability and leverage this bound to design a scalable driver node selection algorithm combining dilation-aware initialization via maximum matching with greedy accessibility expansion. We demonstrate the effectiveness and scalability of the proposed framework through numerical experiments on hypergraphs with tens to thousands of nodes and higher-order interactions.
【28】Minimax Generalized Cross-Entropy
标题：极小极大广义交叉熵
链接：https://arxiv.org/abs/2603.19874
摘要：Loss functions play a central role in supervised classification. Cross-entropy (CE) is widely used, whereas the mean absolute error (MAE) loss can offer robustness but is difficult to optimize. Interpolating between the CE and MAE losses, generalized cross-entropy (GCE) has recently been introduced to provide a trade-off between optimization difficulty and robustness. Existing formulations of GCE result in a non-convex optimization over classification margins that is prone to underfitting, leading to poor performances with complex datasets. In this paper, we propose a minimax formulation of generalized cross-entropy (MGCE) that results in a convex optimization over classification margins. Moreover, we show that MGCEs can provide an upper bound on the classification error. The proposed bilevel convex optimization can be efficiently implemented using stochastic gradient computed via implicit differentiation. Using benchmark datasets, we show that MGCE achieves strong accuracy, faster convergence, and better calibration, especially in the presence of label noise.
【29】On the role of memorization in learned priors for geophysical inverse problems
标题：论记忆在地球物理反问题习得先验中的作用
链接
：https://arxiv.org/abs/2603.19629
摘要：Learned priors based on deep generative models offer data-driven regularization for seismic inversion, but training them requires a dataset of representative subsurface models -- a resource that is inherently scarce in geoscience applications. Since the training objective of most generative models can be cast as maximum likelihood on a finite dataset, any such model risks converging to the empirical distribution -- effectively memorizing the training examples rather than learning the underlying geological distribution. We show that the posterior under such a memorized prior reduces to a reweighted empirical distribution -- i.e., a likelihood-weighted lookup among the stored training examples. For diffusion models specifically, memorization yields a Gaussian mixture prior in closed form, and linearizing the forward operator around each training example gives a Gaussian mixture posterior whose components have widths and shifts governed by the local Jacobian. We validate these predictions on a stylized inverse problem and demonstrate the consequences of memorization through diffusion posterior sampling for full waveform inversion.
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/194277
本文地址：http://www.python88.com/topic/194277