标题：Who Verifies the Verifier? Co-Evolving Inspectable Graders with Self-Improving Agents
来源：arXiv
链接：https://arxiv.org/abs/2610.11464
摘要：We changed the agent: did it actually get better? Every self-improving agent loop answers this hundreds of times, and every answer comes from a verifier. On open-ended tasks none exists, so the loop is handed a hand-written rubric or a bare LLM judge grading output from a model like itself, inviting reward hacking and shared blind spots. We make the verifier the evolving object: an inspectable expression over small, mostly deterministic drawback detectors, synthesized from clustered failures, gated at birth, and selected for agreement with a ten-item anchored reference set plus consensus over unlabeled outputs, never for the agent's score. On MBPP+ it gains +0.21 held-out agreement over the hand-authored seed composition, on every seed, and ends ahead of the bare LLM judge it contains. One
标题：Harness Evolution Hits a Ceiling: When Weight Training Should Begin
来源：arXiv
链接：https://arxiv.org/abs/2610.11655
摘要：Improving a long-horizon LLM agent means evolving the harness around a frozen model or training its weights. We let a self-evolving harness make the system stronger first, then cross seed and evolved harnesses with base and trained weights to learn which gains the trained model keeps and which still need the runtime. We show that the right lever can be read off the agent's failure composition: labelling failed trajectories by the first signal that fires separates process failures (blocked calls, loops, exhausted step budgets) from content failures (a delivered plan that is poor). Harness evolution repairs the former, the behaviour it instils can be trained into the weights, and content failures are what weight training is for. On DeepPlanning, a self-evolving harness loop lifts the held-ou
AgentEvolver 将操作、方法、智能体、控制流、接口和支持状态等八类实体纳入统一的版本化演化生命周期，无需修改基础模型权重。共享 Runtime 负责协调持续任务，持久化规划与可恢复上下文则用于维持任务方向、证据和执行状态。系统在 SWE-bench Pro Public 上取得 82.08% 的问题解决率，优于论文报告的无演化基线。六个应用案例显示，执行中形成的能力可以复用于后续网站、游戏和科研任务，但最终任务成功不等同于可复用组件确实得到改进，跨独立任务迁移与总体开发成本仍待验证。
用途与启示
为冻结基础模型条件下的智能体持续改进提供系统级实现路径。
说明能力演化需要把组件修改、评估、版本管理和后续调用连接成闭环。
提示评估应区分最终任务成功率与可复用能力是否真正积累。
可用于研究智能体执行经验复用、运行时演化及跨任务能力迁移。
📝 原始笔记（逐字保留）
标题：AgentEvolver: System-Wide Self-Evolution Through Task Execution
来源：arXiv
链接：https://arxiv.org/abs/2610.11613
摘要：An agent can complete a task without improving how it works. Turning task experience into reusable capability requires connecting the changed component to its evaluation and subsequent use. We present AgentEvolver, a system for developing capabilities during task execution while keeping the foundation model fixed. Eight entity families expose reusable operations, methods, agents, control flow, interfaces, and supporting state to revision through a common versioned lifecycle. A shared Runtime coordinates ongoing work, while persistent planning and recoverable context preserve task direction and supporting evidence. We evaluate task outcomes on SWE-bench Pro Public and examine capability changes in six application cases. The team reports an 82.08\% resolution rate with evolution, exceeding i
生成式对抗循环：自动设定基准并发现算法
GAL 通过对抗数据生成与算法发现两个智能体搜索过程交替迭代，使系统能够自主暴露弱点、更新评测目标并持续改进算法。
标题：Generative Adversarial Loops
来源：arXiv
链接：https://arxiv.org/abs/2610.11458
摘要：AI research progress can be viewed as the interaction between two processes: benchmark creation and method discovery. Historically, both were driven by human intelligence. However, recent advances in AI have accelerated automated method discovery, while automated benchmark creation has received comparatively less attention. To enable self-advancing systems, we propose Generative Adversarial Loop (GAL), a generator-discriminator framework alternating between two agentic searches: (1) a discriminator that generates adversarial data to expose weaknesses in current systems, and (2) a generator that discovers algorithms to overcome them. We apply this framework to approximation algorithms for efficient inference. Unlike existing auto research systems, which primarily focus on algorithm discover
标题：SAIL: Scientific Agentic Intelligence via a Science-Aware Loop
来源：arXiv
链接：https://arxiv.org/abs/2610.11451
摘要：We introduce SAIL, an open model with 35B total and 3B active parameters for literature research, scientific coding, and multi-step research workflows. SAIL is developed through a science-aware improvement loop: agents built on frontier AI models analyze its task failures and construct training tasks that address the underlying capability gaps. The diagnosis examines search and evidence selection in literature tasks, scientific assumptions and reasoning in coding, and planning and revision in longer investigations. The agents draw on paper collections and scientific code repositories to build problems, interaction trajectories, and executable tasks with the required environments and tools. We repeat this loop over multiple development cycles and train SAIL through supervised fine-tuning, s
标题：How to post-train on a surrogate: Envelope sampling mitigates reward hacking
来源：arXiv
链接：https://arxiv.org/abs/2610.11281
摘要：Large language models (LLMs) are commonly post-trained against LLM judges and other cheap surrogates because the true reward, such as human preference, is too expensive to query at scale. This practice often leads to reward hacking, where reinforcement learning against a miscalibrated surrogate leads to undesirable side effects. In this work, we study a setting in which a small number $n$ of model outputs are annotated with ground-truth labels (e.g., from expert review) and used to recalibrate the LLM judge before optimizing against it. Prior approaches to judge recalibration are costly or heuristic, and it is known that on-policy sampling fails when the surrogate is miscalibrated on a rare set of outputs. In this work, we propose envelope sampling, a theoretically-grounded method for judg
标题：Memory Type Varies: Empowering LLM Agents for Long-Term Memory with Diverse Strategies
来源：arXiv
链接：https://arxiv.org/abs/2610.11573
摘要：The memory capabilities of Large Language Models (LLMs) have garnered increasing attention recently. Despite great success achieved, existing retrieval-based memory approaches typically overlook the differences between memories and employ a unified strategy to process all memories, leading to suboptimal performance. Thus, an intuitive question arises: can we categorize memory into different types and select appropriate strategies? However, given the topic-rich, scenario-complex, and boundary-blurred nature of memory scenarios, achieving precise classification of memories is not easy. To address this challenge, we propose a memory multi-class dataset in this paper, termed TriMEM, which provides precise annotations for memory types across diverse scenarios. Building upon this foundation, we
标题：ReTeach: Building a Self-Teacher through Multi-Round Reflection and Retry
来源：arXiv
链接：https://arxiv.org/abs/2610.11529
摘要：Self-distillation can improve reasoning without a separately trained, more capable teacher, but its effectiveness depends on how the self-teacher gains an advantage over the student. Conditioning the teacher on reference answers or solutions can provide such an advantage, but this information may be unavailable. Reflection offers a way to derive explicit error diagnoses and revision guidance from self-generated attempts, yet existing reflection-based methods often combine it with reference information, rich task feedback, or persistent memory. We introduce ReTeach, a Reflective self-distillation framework that constructs its self-Teacher through multi-round reflection and retry using only self-generated attempts and outcome-level verification. Starting from an unsuccessful student rollout,
标题：SynCo: Data Synthesis Co-Training for Self-Evolving LLMs via Multi-Agent Reinforcement Learning
来源：arXiv
链接：https://arxiv.org/abs/2610.11345
摘要：Self-evolving LLM agents promise to improve autonomously through continual interaction and learning, reducing their dependence on manually curated supervision. Realizing this promise requires not only updating the agent, but also evolving its training experience as its capabilities change. However, most existing pipelines rely on static datasets or separately updated synthesis models, causing previously useful tasks to become trivial while overly difficult tasks remain uninformative. This growing mismatch between agent capability and training experience limits sustained self-improvement. To address this problem, we propose SynCo, an agentic data synthesis co-training framework for self-evolving LLMs based on multi-agent reinforcement learning. SynCo jointly optimizes two independently para
标题：Adversarial Cues in Decision Models Used as Judges: The Role of Request Presentation
来源：arXiv
链接：https://arxiv.org/abs/2610.11436
摘要：An answer judge instructed to grade the final commitment should reject an explicitly wrong final value even when an earlier value matches the reference. We show that adding one colon to a candidate can violate this requirement depending on the presentation of the structured judging request. Numeric references certify the error, and paired interventions distinguish the candidate edit from the integration's presentation choices. On 200 previously unused DROP and GSM8K source clusters, the edit increased Jev's false acceptance from 1.0% to 26.0% with three output labels and from 3.0% to 26.5% with the published four-label grading instruction under sorted JSON keys. Both candidate variants were rejected under insertion presentation. These interactions passed the prespecified statistical correc
TypedBench：System One 决策模型的校准、措辞敏感性与成本评测
TypedBench 从策略遵循、措辞鲁棒性、概率质量和实际决策成本等维度，系统评测输出类型化概率的 System One 决策模型。
标题：TypedBench: A Benchmark for Calibration, Framing Sensitivity, and Cost in System One Decision Models
来源：arXiv
链接：https://arxiv.org/abs/2610.11392
摘要：System One models output calibrated probabilities over typed answers such as categorical choices, ordinal levels, or binary outcomes, via a non-generative interface. Software can act on these probabilities through thresholds, cost-weighted choices, and escalation rules. Consequently, if these probabilities are miscalibrated or wording-sensitive, the software ma take unintended actions leaving human operators with no textual rationale to inspect. Current evaluations largely report accuracy and calibration on public classification datasets without a clear reference. We present TypedBench, a benchmark for typed decision models like Jev built from seven policy-labelled generators and nine evaluation suites. We report accuracy as median and range across paraphrases, and calibration error relati
企业Agent上线后会持续面对界面、流程和用户偏好的变化，竞争焦点因此从单次任务完成转向能否利用失败轨迹与业务反馈持续改进。产业化RSI可拆为基础模型智能、Agent Harness以及后训练三层，其中企业场景尤其依赖任务编排、Reward、Verifier、评估和安全发布机制。Pyromind提出以AutoRL和Reward Data Flywheel串联真实任务轨迹、奖励生成、策略更新、模型评估与部署反馈，并通过软件Agent、工业规则和具身数据生成案例验证跨场景复用能力。文章强调，RSI基础设施的核心壁垒不是单点指标提升，而是能否将客户现场经验抽象为可归因、可迁移、可审计的持续学习机制，同时防范过拟合、奖励作弊和能力退化。
DreamTrue是一个支持多视角、跨本体的机器人世界模型，目标是根据动作生成忠实且符合物理规律的未来视频。该方法将动作轨迹渲染为图像空间条件，并通过离线几何校准缓解机器人数据标定误差造成的动作错位。其反事实后训练会修改已记录的动作轨迹，从而覆盖更多动作与接触配置，尤其补充真实数据中稀缺的失败交互。团队还利用人工标注的机器人、物体及交互缺陷视频训练具身视频奖励模型，并以其反馈开展强化学习后训练；在AgiBot上，交互缺陷率由48.12%降至6.25%，并获AgiBot World Challenge 2026世界模型赛道第一名。
标题：Conditional Transfer from Controlled Pretraining Mixtures to Code
来源：arXiv
链接：https://arxiv.org/abs/2610.11548
摘要：Synthetic tasks are increasingly used both as probes of language-model capability and as pretraining data. Both uses are often justified by loss reduction: falling loss is treated as informative, and faster loss reduction with more sampling as evidence that a task is worth sampling. We separate three signals. A task is diagnostic when its loss tracks global pretraining progress; it is teachable when its loss responds to its own token budget; and a data source transfers when including it improves a downstream target. We study controlled pretraining in which 70% of the corpus is fixed general Python and the remaining 30% is a simplex over three source families: OpenCodeInstruct, a curated suite of 12 code-adjacent synthetic tasks, and 15 literature-derived probe tasks. Across a task-budget s
标题：SpatialOPSD: Self-Distilling Spatial Intelligence from Verified Coding Agent Traces
来源：arXiv
链接：https://arxiv.org/abs/2610.11366
摘要：Spatial coding agents significantly improve spatial reasoning in Multimodal Large Language Models (MLLMs) by using external tools to generate verified execution traces. However, this paradigm inherently suffers from prohibitive inference-time overhead and external dependencies. In this paper, we explore whether an MLLM can internalize this agentic capability to operate entirely tool-free. We begin with a simple observation: prompting an MLLM with summarized execution traces of a spatial coding agent naturally unlocks the model's internal spatial Chain-of-Thought (CoT). Motivated by this, we introduce SpatialOPSD, an on-policy self-distillation framework that internalizes spatial reasoning into a standalone MLLM by formulating verified agent traces as privileged information. To mitigate pri
标题：UniData: Universal Multimodal Instruction Generation Pipeline
来源：arXiv
链接：https://arxiv.org/abs/2610.11363
摘要：Multimodal Large Language Models (MLLMs) are increasingly being applied in a wider range of real-world scenarios. However, due to the substantial labor cost, creating high-quality multimodal instruction datasets for MLLMs remains a significant challenge. Although some methods propose to generate instruction data, they often face limitations in modality support and struggle with generating multi-round instructions. To address these problems, we introduce UniData, a universal instruction generation pipeline, to transform simple user requirements into multi-round, multimodal instructions. Specifically, UniData first expands user requirements into multiple diverse events. Using these events, UniData then integrates an any-to-any large model for multimodal instruction generation. Finally, UniDa
标题：SWE-Journey: Towards More Realistic Evaluation of Coding Assistants through Long-Horizon, Multi-Turn Interaction
来源：arXiv
链接：https://arxiv.org/abs/2610.11559
摘要：Coding assistants such as Claude Code and Codex have become a major application of LLM agents, yet existing benchmarks remain far from real-world use, particularly in task horizon and interaction length. Code assistants require completing long chains of development work in continuously evolving repositories, while repeatedly clarifying requirements and adapting implementations through multi-turn interaction. To address these gaps, we introduce SWE-Journey, a benchmark for more realistic evaluation of coding assistants. To address the task-horizon gap, we propose a weak-to-strong synthesis pipeline that automatically constructs long-horizon coding tasks. To address the interaction gap, we mine four representative user personas from real interaction data and build a user-simulation agent to
标题：MedBenchAgent: Towards Systematic Automation of Medical VLM Benchmark Construction
来源：arXiv
链接：https://arxiv.org/abs/2610.11312
摘要：Large-scale construction of medical vision-language model (VLM) benchmarks is increasingly feasible with richly annotated imaging datasets and large language models (LLMs), yet existing automation largely focuses on generating evaluation items within predefined benchmark specifications. We study the broader problem of automatically deriving the specification itself: what to evaluate, which annotations support each task, and how to translate this evidence into reliable evaluation items. We formulate benchmark construction as constrained compilation, in which the benchmark specification is progressively derived from evaluation requirements, heterogeneous annotations, and medical knowledge. Based on this formulation, we introduce MedBenchAgent, a multi-agent framework with a Benchmark Interme
标题：Error-Propagation Modeling for Failure Attribution in LLM-Based Multi-Agent Systems
来源：arXiv
链接：https://arxiv.org/abs/2610.11600
摘要：LLM-based multi-agent systems (MASs) are increasingly used to solve complex tasks through coordinated reasoning, tool use, and interaction with external resources. However, attributing failures in such systems remains challenging because the observed outcome often does not directly reveal the error responsible for the failed execution. In this work, the attribution target is the decisive error, defined as the agent--step pair whose correction would recover the failed execution. Existing approaches largely identify suspicious steps without explicitly modeling how errors propagate across interactions or persist in unresolved loops, making decisive errors difficult to distinguish from downstream failure symptoms. We propose \textbf{E}rror-Propagation \textbf{M}odeling for \textbf{F}ailure \te
★2模型迭代削弱 AI 辅助科学写作检测可靠性arXiv · 评测审计 · 论文对23个版本模型生成的科学摘要进行跨版本测试发现，写作检测器会在模型代际更替时性能骤降，固定基准上的准确率不足以长期支撑科研诚信审查。
标题：Large Language Model Turnover Undermines Screening for Artificial Intelligence-Assisted Scientific Writing
来源：arXiv
链接：https://arxiv.org/abs/2610.11599
摘要：Journals and conferences have begun to screen submitted manuscripts for text written using large language models (LLMs). The reliability of this screening rests on benchmark evaluations against a fixed set of LLM versions, while the versions in actual use keep changing. Here we quantify how this LLM turnover affects the screening of scientific manuscripts. We paired 4,000 pre-ChatGPT abstracts from the Proceedings of the National Academy of Sciences with their rewrites by 23 LLM versions from three vendors, released between June 2023 and August 2026. We then trained detectors under maintenance scenarios ranging from a detector retrained on every new version to one trained once and never updated. Detectors trained only on a vendor's past versions can collapse at the boundaries between model
E-Ledger 是面向企业工作流的多智能体执行框架，在每项操作执行前通过代码审批层检查其是否符合组织策略。系统使用世界账本持久保存已验证的隐藏规则，以及由证据支持的动态环境状态，以应对部分可观测和跨记录的长程任务。WorldAbduct 从状态一致性、世界—观测差距、策略门控正确性和目标判断四个视角诊断执行轨迹，并通过定向溯因交互验证潜在规则后写入账本。在 World of Workflows 上，该方案跨四种 LLM 主干将安全任务完成率较最强演化基线提高 5—15 个百分点，并在 ScienceWorld 和 DiscoveryWorld 中展现出向科学环境迁移的能力。
用途与启示
为高风险企业工作流提供“先验证、后执行”的策略门控机制，降低智能体违规操作风险。
通过证据支持的持久化世界账本统一管理动态状态与隐藏规则，增强长程任务中的状态一致性。
展示了利用执行失败和观测缺口进行溯因式框架演化的路径，可用于持续发现并验证环境机制。
可作为智能体可信执行、轨迹诊断和世界模型驱动脚手架演化的参考架构。
📝 原始笔记（逐字保留）
标题：Safe, Persistent, and Evolving Agent Harness for Understanding Partially Observable Worlds
来源：arXiv
链接：https://arxiv.org/abs/2610.11552
摘要：Large language model agents can invoke tools fluently, but enterprise workflows demand more than selecting the right tools: actions must strictly comply with organizational policies, tool feedback often conceals hidden side effects under partial observability, and long-horizon tasks require persistent state tracking across multiple records. To address these challenges, we introduce E-Ledger, a multi-agent harness for safe and persistent execution. E-Ledger employs a code approval layer that checks every proposed action against policy before execution, and maintains a world ledger of verified hidden rules alongside evidence-backed dynamic state. Because hidden rules are typically unknown a priori, we further propose WorldAbduct, an abductive, world-model-driven harness evolution framework.
标题：DuplexAgent-RSI: Recursive Harness Improvement for Full-Duplex Voice Agent Collaboration
来源：arXiv
链接：https://arxiv.org/abs/2610.11299
摘要：Voice agents are converging on a collaboration pattern: a full-duplex interaction model stays on the live channel as the entry to the conversation, while search, reasoning, and coding are handled through asynchronous delegation. A duplex model supports continuous listening and speaking, but complex reasoning and tool use may exceed its capabilities. A coding agent can plan and execute extended tasks, but its sequential interface is a poor fit for live conversation. Combining them requires a harness that coordinates task acceptance, progress, cancellation, replacement, and result delivery while keeping the conversation responsive. Existing harnesses often rely on coupled heuristics, making them difficult to improve systematically from evidence. We present DuplexAgent, a full-duplex collabor
标题：Harness Compilation: Which Decisions Should a Small Vision-Language Model Keep?
来源：arXiv
链接：https://arxiv.org/abs/2610.11231
摘要：Small vision-language models may be able to read external evidence yet struggle to obtain it. We introduce Harness Compilation (HC), an offline procedure that adapts the division of work between a frozen small VLM and its external harness. A large teacher uses student execution traces to revise reusable content and control, while a separate validation set selects the deployed harness. Deployment requires neither weight updates nor teacher calls. Across seven visual question-answering settings with students of at most 9B parameters, HC improves scores over bare students by 9.9-23.9 points, averaged over three independent builds per setting. Interventions on five runtime decision types (invocation, selection, argument generation, evidence integration and abstention) show why this allocation
标题：One Skill Too Many: How Co-Installed Skills Conflict in Coding Agents
来源：arXiv
链接：https://arxiv.org/abs/2610.11647
摘要：Coding agents are extended with agent skills, directories whose SKILL.md tells the model when and how to perform a task. Because skills come from independent sources (teams, developers, plugins, copied collections), an installed skill can be co-installed with a similar skill doing the same job, and the model picks between them by name and description alone. In a conflict, the installed skill loses core functions (e.g., a ban on touching git) because the similar skill runs instead or changes what it does. The task still passes, so benchmarks that check only task completion miss such cases. We present the first empirical study of such conflicts. From snapshots of 20,947 repositories, we mine 822,109 candidate similar-skill pairs, have an LLM judge a stratified sample of 3,754, and run 312 co
标题：Constitutional Gating and Deterministic Recovery for Multi-Agent LLM Negotiation: Ablations Against a Stateful Adversarial Gatekeeper
来源：arXiv
链接：https://arxiv.org/abs/2610.11542
摘要：Multi-agent LLM systems negotiating with a stateful counterpart waste model calls in three ways: polite loops that never meet the counterpart's hidden acceptance condition, malformed outputs that trigger retries, and compliance deadlocks in which the counterpart demands something the agent must refuse. We study a three-part control stack - a 5-Pillar runtime constitution, a 4-tier swarm (Director, three-agent majority vote, Monitor, schema hard gate) and Cognitive Annealing (deterministic deadlock detection, atomic purge of the agent-side context, a canonical recovery message) - against a released adversarial Gatekeeper whose acceptance rules are fixed regular expressions and whose LLM only renders reply text. The testbed has a known solution: it measures whether the stack executes a const
标题：Evaluating Local Language Model Agents for Reproducible Data Engineering: An Empirical Software Engineering Study of Mobility Workflows
来源：arXiv
链接：https://arxiv.org/abs/2610.11482
摘要：Context: Large language model (LLM) agents are increasingly used as software and data-engineering assistants, yet evidence about locally deployable open-weight agents remains limited. Existing evaluations often emphasize textual responses or isolated code generation rather than the validity of complete engineering artifacts. Objectives: We evaluate whether local LLM agents can produce correct and reproducible data-engineering artifacts, quantify the effect of a closed-loop workspace condition, and examine trade-offs in model scale, architecture, quantization, runtime, tool use, and failure. Methods: We introduce a benchmark of fifteen mobility-workflow tasks covering data discovery, connectors, transport-feed processing, semantic enrichment, feature engineering, validation, visualization,
标题：Closed-loop evaluation of LLM agents for embedded software development
来源：arXiv
链接：https://arxiv.org/abs/2610.11447
摘要：Large language models (LLMs) are increasingly deployed as coding agents that edit files, run builds and tests, inspect execution results, and repair software iteratively. Embedded firmware is a demanding target because correctness depends on closed-loop behavior under sensing, timing, and safety constraints, not only on static source quality. Yet embedded-agent evaluation remains limited and often emphasizes one-shot synthesis or offline correctness. We present a benchmark for closed-loop evaluation of embedded coding agents. Each task provides a plain-text engineering description, constrained workspace, and visible build-and-runtime surface. The agent must translate requirements into implementation and self-verification steps, then iterate until the required device behavior is achieved. T
MetaEncoder 面向输出受约束决策及概率分布的 System One 场景，以自然语言替代结构化 schema 来描述用户请求和候选选项。该方法将预训练的 Muse-Glimmer 30B 解码器微调为遵循指令的决策编码器，并可接收图像和视频等辅助输入。系统采用双编码器架构与单向对比学习进行请求—候选对齐，可同时覆盖少于 256 个候选的闭集任务和百万级候选的开放集检索。作者在 11 组基准、190 项任务上验证其多模态决策、理解与检索能力，同时指出其在推理密集型任务上仍有局限。
用途与启示
为大规模候选空间中的低延迟决策与检索提供可扩展的多模态双编码器方案。
展示自然语言接口可以统一状态、意图和候选项表示，降低结构化 schema 的设计与维护成本。
可用于推荐、搜索、内容选择等需要输出约束决策或概率分布的应用。
其推理密集型任务表现提示，快速 System One 编码器仍需与更强的显式推理机制协同。
📝 原始笔记（逐字保留）
标题：MetaEncoder: Exploring the Limit of Bi-Encoders for Multimodal System One Decision Making with Natural Language Interface
来源：arXiv
链接：https://arxiv.org/abs/2610.11316
摘要：System One models output constrained decisions and probability distributions rather than free-form text generation. While prevailing paradigms rely on structured schema objects to encode state, intent, and candidate choices, we revisit a fully natural language-based System One interface. In this framework, both the user request and each candidate option are expressed in natural language, supported by multimodal (image and video) auxiliary inputs. We introduce MetaEncoder, which fine-tunes a pre-trained Muse-Glimmer 30B decoder into an instruction-following decision-making encoder. To scale effectively across both small closed-set (< 256) and massive open-set (millions) candidate spaces, MetaEncoder employs a bi-encoder architecture trained via unidirectional contrastive learning for reques
标题：Project Greenhouse: Progress Toward Fully Open and Sovereign Agentic Search
来源：arXiv
链接：https://arxiv.org/abs/2610.11922
摘要：Project Greenhouse represents our exploration of a simple thesis: We believe that it is possible to build fully open and sovereign models for agentic search with only modest computational resources. As a first milestone, we describe how to build a competitive pointwise decoder-only reranker using a simple two-step recipe comprising pre-training from scratch followed by supervised fine-tuning, starting only from commonly available datasets. Contrary to the dominant approach in the literature, we do not rely on existing open-weight backbones from third parties, and thus we are fully in control of model training, from end to end. We were able to accomplish the bulk of our experiments using no more than a handful of GPUs. This report articulates the importance and benefits of our approach, and
标题：RoboAware: Learning to Coordinate Embodied Skills from Counterfactual Outcomes
来源：arXiv
链接：https://arxiv.org/abs/2610.11480
摘要：Embodied coding agents can combine modular robot skills with frozen end-to-end policies, yet effective composition requires anticipating which policy family will succeed in the current physical state. We present RoboAware, which builds on coding agents' skill orchestration by learning only a state-conditioned responsibility coordinator from counterfactual outcomes. Inspired by the success of REPL, we propose the $P^5$ schema and formulate a hierarchical MDP based on it. $P^5$ organizes skills uniformly into five semantic stages, defining where responsibility can be compared. To address the lack of counterfactual branch outcomes in existing work, we introduce State-Locked Counterfactual Branching (SCB), which restores the same training state to generate and execute a code block from each ad
标题：An Interpretable Approach to PDE Solution Discovery via Structural Experience Distillation
来源：arXiv
链接：https://arxiv.org/abs/2610.12003
摘要：PDE solution discovery aims to identify explicit symbolic expressions for unknown physical fields from observations under known physical constraints. Existing methods, however, collapse data fidelity and physical consistency into a single terminal score used as the sole feedback signal, providing little information about which subexpressions are responsible for a candidate's final performance. This opaque terminal feedback severely limits the interpretability of the search process itself, offering no insight into why a candidate succeeds or fails. Consequently, reusable structures in otherwise suboptimal candidates are often discarded, whereas incidental syntax along successful search trajectories may be repeatedly reinforced. We propose SED-MCTS, a Monte Carlo tree search approach that di
标题：EvoSim: Learning to Model, Modeling to Learn
来源：arXiv
链接：https://arxiv.org/abs/2610.11344
摘要：Physics-based models connect scientific explanation with quantitative prediction. Constructing them requires selecting physical processes, defining states and governing equations, specifying couplings, and identifying parameters from experiments. Existing AI systems remain limited in making these model structure decisions autonomously. We introduce EvoSim, a self-evolving AI scientist for physical modeling. It uses experimental discrepancies to drive mechanism and equation revisions and held-out experimental data to test physical plausibility. Exploration traces make updates to knowledge, skills, and multi-agent orchestration. This co-evolution improves physics-based models and EvoSim's ability to select mechanisms, diagnose failures, and coordinate research. We evaluate EvoSim on two indu
标题：Open-ended Scientific Discovery with Possibilistic Reasoning
来源：arXiv
链接：https://arxiv.org/abs/2610.11289
摘要：Autonomous scientific discovery with LLMs requires generating and testing hypotheses adaptively as evidence accumulates while maintaining statistical validity. Existing anytime-valid methods can handle data-dependent hypotheses, but open-ended discovery poses a deeper challenge: the best discovered hypothesis may still be the best of a bad lot, with better explanations yet undiscovered, while even background knowledge such as physical laws may require revision in light of new findings. In response, we formalize the problem as Abductive Autonomous Scientific Discovery (AASD) using possibility theory. We introduce abductive utility, a computable measure of discovery progress, and possibility frontier search, the first algorithm for AASD, which maintains anytime validity and achieves $\vareps
标题：LLM-IDEA: Identifiability-Driven Experimental Agent for Autonomous Discovery of Mechanistic World Models
来源：arXiv
链接：https://arxiv.org/abs/2610.11253
摘要：Large language model agents are being increasingly deployed as autonomous scientists, designing experiments and inferring mechanistic world models with minimal human oversight. Yet identifiability is often overlooked: when a plateau is reached, the agent needs to know whether it is not yet capable enough or the model simply is not identifiable from the data, in which case no amount of further experimentation of the same kind can help. We propose the Identifiability-Driven Experimental Agent (LLM-IDEA) for closed-loop discovery with an identifiability engine that returns a three-way plateau verdict: capability limit, resolvable within the design class, or certified exhausted. On ODEBench, 60 of the 62 systems with free constants are identifiable at round 0; the RC circuit is certified exhau
标题：RISR: Residual-Informed Scientific Equation Discovery with Large Language Models
来源：arXiv
链接：https://arxiv.org/abs/2610.11387
摘要：Symbolic regression combines structural search with numerical fitting, but aggregate fit scores do not describe how the remaining error varies across inputs. We introduce RISR, a residual-informed method that uses these error patterns to guide formula discovery and learn which corrections are worth fitting. A residual encoder compresses aligned inputs, targets, current predictions, and residuals into continuous tokens that condition a language model to propose formulas. For subsequent refinement, a dual-view relational encoder uses additive and regularized multiplicative residuals to predict the post-fit utility of candidate corrections. We evaluate RISR on scientific tasks from the LLM-SRBench. RISR achieves 63.57% and 38.50% ID accuracy at the 1% and 0.1% pointwise relative-error toleran
论文提出 Incremental-OEDR，使开放式深度研究系统无需在新信息出现时从头生成报告，而是保留有效知识、修订过时内容并增量补充证据。Structured Harness 将报告表示为大纲、章节和支撑证据的结构化集合，并结合结构化检索、持久化证据池与选择性生成实现证据复用。作者建立了跨度十年的时序评估框架，通过 Single-Step Task 和 Long-Chain Task 分别衡量单次更新与长期连续更新能力。在 DeepResearch Bench 上，该方法最高将内容级 ROUGE-L F1 提升 0.51、大纲级 EM F1 提升 0.63，同时减少 33% 的词元消耗和 61% 的搜索调用。
用途与启示
为需要长期维护的研究报告提供增量更新范式，避免重复执行完整调研流程。
通过结构化证据池保留来源与支撑关系，有助于提升报告连续性、证据复用率和更新可控性。
时序长链评估可用于检验深度研究智能体在知识过时、内容修订和长期状态维护方面的能力。
选择性检索与生成机制可显著降低搜索调用和词元成本，适合持续情报跟踪与动态知识报告。
📝 原始笔记（逐字保留）
标题：Incremental Open-Ended Deep Research with Structured Harness
来源：arXiv
链接：https://arxiv.org/abs/2610.11566
摘要：Existing Open-Ended Deep Research (OEDR) systems primarily generate reports from scratch, making them inefficient for scenarios where research reports need to be continuously maintained as new information emerges. We introduce \textbf{Incremental Open-Ended Deep Research (Incremental-OEDR)}, a research setting that treats a report as an evolving research state and incrementally updates it by preserving valid knowledge, revising outdated or incomplete content, and incorporating newly available information. To support this setting, we propose \textbf{Structured Harness}, which represents reports as structured collections of outlines, sections, and supporting evidence, and provides structured retrieval, a persistent structured evidence pool, and structured generation for selective report upda
标题：Adapting English Quality Classifiers for Multilingual LLM Pretraining Data Selection
来源：arXiv
链接：https://arxiv.org/abs/2610.11585
摘要：Recent advances in large language model (LLM) pretraining highlight the role of high-quality training data in improving performance. While model-based filtering has proven effective in selecting high-quality subsets from web-scale corpora, especially for high-resource languages, low-resource languages face challenges due to limited availability of annotated data. This work explores extending quality filtering to over 100 languages by proposing a multilingual adaptation approach that converts an existing English quality classifier into a multilingual variant. Our approach proposes training a small multi-layer perceptron on top of Transformer encoder-only model embeddings, using multilingual text as input and scores obtained from English classifiers applied to machine-translated text as labe
标题：ReCast: Attribution-Oriented Step Representation Learning for LLM-Based Agent Systems
来源：arXiv
链接：https://arxiv.org/abs/2610.11334
摘要：In LLM-based agent systems, failures can originate from early steps whose effects propagate through subsequent interactions, making their origins difficult to identify. To trace such failures back to their origin, failure attribution has been formulated as the task of identifying the earliest step responsible for the failure. Recent methods leverage LLM internal signals for failure attribution, typically using hidden states as step representations. We therefore conduct an empirical study to evaluate how effectively these representations distinguish root-cause steps from other steps and find limited separation. Motivated by this observation, we propose ReCast, a step representation learning method that transforms hidden states from a frozen LLM into attribution-oriented step representations
标题：From Retrieval to Reconstruction: Constructing Evolvable Cognitive Memory for Long-Term Dialogue
来源：arXiv
链接：https://arxiv.org/abs/2610.11314
摘要：Large Language Models (LLMs) serving as long-term dialogue agents require memory systems that support reliable reasoning over extended interactions. However, existing Retrieval-Augmented Generation (RAG) frameworks typically treat memory as passive storage, making it difficult to distinguish source-attributed beliefs from unattributed event/fact records and to connect evidence dispersed across sessions. We introduce CogMem, a cognitive memory architecture based on the PEC$^2$F (Person-Event-Concept-Claim-Fact) graph schema. Dedicated Claim nodes preserve the source and target of subjective statements, while Fact and Event nodes represent semantic and episodic knowledge. Dialogue turns are incrementally converted into provenance-aware graph records, consolidated into higher-level facts, and
标题：Gated Memory: Admission-Controlled Memory Formation for Conversational AI
来源：arXiv
链接：https://arxiv.org/abs/2610.11270
摘要：Personalized conversational AI relies on long-term memory systems that extract facts from user utterances and store them in persistent vector stores. Despite progress in retrieval, deduplication, and lifecycle management, the formation stage, the moment a fact is first written to storage has received almost no principled attention. We identify this as the binding constraint on memory quality in production systems. Critical contextual signals, such as the distinction between a permanent user attribute and a transient situation, exist only in the original utterance and are irreversibly lost the moment extraction produces a subject-relation-object triple. No downstream process can recover them. We propose Gated Memory, a lightweight, modular formation framework that interposes two decision ch
标题：What to Admit and How to Present: Governing Persistent Memory in LLM Agents
来源：arXiv
链接：https://arxiv.org/abs/2610.11188
摘要：Persistent memory can improve personalization in LLM agents but can also induce sycophancy and cross-domain leakage. We distinguish two governance decisions: admission, which determines what recalled information enters the working context, and presentation, which determines how admitted information is expressed. We implement two inference-time designs without retraining: factor-compiled admission (FC), which assesses whole memory entries, and permission-semantic admission (PS), which decomposes entries into typed units; both translate adjudicated attributes into eligibility decisions via deterministic policies. We evaluate on a four-backbone development suite and an external benchmark with four tasks of 300 samples each. Relative to verbatim injection, FC and PS reduce pooled judge-assesse
MARI 针对现有推荐方法将浏览、购买等异构行为扁平化为同质序列而造成决策语义丢失的问题，引入显式的结构化决策证据。系统通过事后决策蒸馏，从异构行为和用户生成内容中离线提取目标、约束及权衡，形成结构化决策记忆并存入决策记忆库。在线推理时，MARI 检索相关记忆以增强大模型推荐，无需反复处理冗长的原始行为序列。实验显示，该方法在标准下一物品预测和新提出的困难选择预测任务上优于现有基线，同时仅带来较低的在线延迟开销。
用途与启示
为多行为推荐提供一种保留行为角色、目标约束和决策权衡的结构化用户记忆表示。
通过离线记忆构建与在线检索解耦，降低长行为序列带来的推理成本和延迟。
适用于高度相似候选项之间的困难选择，并能输出可读的用户决策依据。
可为推荐系统中的长期偏好建模、经验检索和可解释推理提供设计参考。
📝 原始笔记（逐字保留）
标题：Beyond Sequences: Distilling Structured Decision Memory for LLM Recommendation
来源：arXiv
链接：https://arxiv.org/abs/2610.11501
摘要：Despite the adoption of large language models (LLMs) in recommendation systems, prevailing approaches mostly model single-type behaviors (e.g., views or purchases). Even when incorporating multiple behaviors, existing methods flatten heterogeneous actions into homogeneous token sequences, ignoring their distinct decision-making roles. This flattening fails to capture semantic hierarchies and contextual nuances in complex decision-making, such as trade-offs between price and quality. Consequently, performance degrades in critical ``difficult-choice'' scenarios involving highly similar items. To bridge this gap, we propose MARI (Memory-Augmented Recommendation with Interpretability), which grounds predictions in explicit, structured decision evidence. MARI maintains a Decision Memory Bank (D
标题：REMORY: Learning Residual Memory for Context Compaction
来源：arXiv
链接：https://arxiv.org/abs/2610.11287
摘要：Long-horizon agents compact their history to continue within a finite context window, but a textual summary alone may not support every subsequent decision. We introduce REMORY, a neural memory network that supplements the summary with a bounded sequence of soft memory tokens. Given the history and summary, the network learns to generate tokens that help a frozen LLM approximate the continuation it would produce with the full history. The tokens are conditioned on the summary and appended after it, forming an analogue of a residual connection along the sequence dimension. On SummHay, REMORY improves source attribution at nearly unchanged insight coverage and approaches the full-context joint score using only 5.2% of the input positions. Across long-horizon agent benchmarks, Qwen3.8-27B and
标题：Traceable World State: A Provenance-Aware State Representation and Deterministic Replay Framework for Robotic Systems
来源：arXiv
链接：https://arxiv.org/abs/2610.12033
摘要：Robotic systems operating over extended tasks must maintain a world state assembled from observations arriving at different times, with varying confidence and potential revisions. Conventional representations emphasize latest estimates, hindering fact provenance, decision reproduction, or execution auditing. We present Traceable World State (TWS), a middleware-neutral semantic representation and reference runtime for provenance-aware robot world state. A TWS snapshot captures entities, relations, observations, confidence, and revision metadata. Validated update operations transform snapshots immutably, ordered updates support deterministic replay, and a canonical SHA-256 hash chain ensures tamper-evident logs. We evaluate TWS through schema conformance, complete state lifecycles, determini
标题：SkillContrast: Difference-Guided Text Selection for Agent Skill Reranking
来源：arXiv
链接：https://arxiv.org/abs/2610.11650
摘要：Similar agent skills can share instructions but differ in their conditions of use. Query-based text selection may retain shared instructions and omit these distinctions. We introduce SkillContrast, a training-free selector that compares retrieved skills and retains their differing text with local context for a pretrained reranker. On 1,235 requests from SameCapRisk-Bench, it yields 54-72 more clean hits (requests that retrieve a helpful skill without its marked risky sibling) than TF-IDF query selection at identical per-candidate input lengths, across 2 retrievers and 2 reranker sizes. Length-matched component replacements identify differing text as the main contributor in the primary setting, with smaller, mixed context effects. Relative to full skill bodies, SkillContrast uses 51.1-58.8%
标题：Chronos Enables Code Agents to Reason over Software Evolution
来源：arXiv
链接：https://arxiv.org/abs/2610.11578
摘要：Historical pull requests record the design decisions, compatibility constraints, and implementation patterns behind a codebase's current state. Experience relevant to a new task can span related changes whose descriptions emphasize different concerns. We introduce Chronos, a test-time framework that makes this connected history available to large language model (LLM)-based code agents. Chronos distills merged pull requests into structured experience cards and connects them through a typed graph of code-level, developer-intent, and organizational relations. Semantic search identifies entry cards, and weighted multi-hop expansion retrieves connected changes for selective reading. The same memory guides candidate generation and patch selection: a patch-focused change agent and a validation-st
标题：Tracing the Thoughts of a Coding Agent Playing ARC-AGI-3: Lessons for Continual Learning
来源：arXiv
链接：https://arxiv.org/abs/2610.11450
摘要：We study how a coding agent learns across a sequence of abstract reasoning tasks. The agent runs on a frozen foundation model inside a fixed harness and acts by writing and running Python and shell scripts. It retains no state across turns other than its written artifacts, so every thought it forms, carries, corrects or abandons leaves a trace, where a thought is any belief, rule or plan committed to a file. We let the agent play ARC-AGI-3, a set of interactive reasoning games that provide no instructions. Each game is a sequence of levels, and a strategy that clears one level can fail on the next, so every new level is in effect a new task. The agent records what it learns as Python scripts and text notes, while the harness keeps a complete log of every action and observation. Our contrib
标题：VAMR: Multi-Question Agentic Reasoning for Efficient Long-Form Video Understanding
来源：arXiv
链接：https://arxiv.org/abs/2610.11171
摘要：Long-form video understanding often involves multiple questions about different aspects of the same recording. Yet existing video agents typically process each question through an isolated tool-use trajectory. This repeatedly restarts video exploration and memory construction, missing opportunities to acquire evidence jointly and progressively build a shared understanding that supports the complete question set. We introduce \textbf{VAMR} (\textbf{V}ideo \textbf{A}gent for \textbf{M}ulti-Question \textbf{R}easoning), which coordinates all questions about a video through one shared tool-use trajectory. At each round, a persistent policy model can invoke tools for one or more unresolved questions and submit answers for questions with sufficient evidence. Question-conditioned visual perceptio
标题：Residual Advantage: Student-Relative Teacher Guidance for RL with Verifiable Rewards
来源：arXiv
链接：https://arxiv.org/abs/2610.11519
摘要：Reinforcement learning with verifiable rewards (RLVR) and on-policy distillation (OPD) have become two main paradigms for post-training reasoning models. RLVR gives each response a single outcome label, leaving the steps inside it without separate credit. OPD provides token-level guidance at student-visited prefixes, but its pointwise signal does not directly reflect the pattern of teacher--student disagreement across the vocabulary. Dense, unbounded log-ratio supervision can amplify the teacher's influence, yet a strong solver is not necessarily a suitable guide when the student's solution paths depart from the teacher's. We propose Residual Advantage (\RA{}), which treats the teacher--student probability residual as a bounded one-step reward, subtracts the corresponding state value under
标题：ReCal: Calibrating Structured Pruning for On-Policy Distillation Recovery
来源：arXiv
链接：https://arxiv.org/abs/2610.11332
摘要：Structured pruning reduces the deployment cost of reasoning language models, but the resulting capability degradation can hinder subsequent on-policy distillation (OPD) recovery. Because OPD relies on student-generated trajectories, pruning damage that persists after offline distillation can limit its effectiveness. We propose RECAL, Recovery-Aware Calibration, a simple plug-and-play approach that improves OPD recovery by adjusting calibration before pruning. RECAL uses forward KL between an unpruned teacher and a pruned probe to identify teacher-supported predictions disrupted by pruning, then reweights calibration statistics to guide existing pruning criteria toward preserving these predictions. Across multiple models and pruning methods, RECAL consistently improves mathematical reasonin
标题：When Do We Need On-Policy Distillation? Distilling on Offline Student Rollouts Is Often Better
来源：arXiv
链接：https://arxiv.org/abs/2610.11291
摘要：On-policy distillation (OPD) has become increasingly popular for transferring teacher capabilities to student models. In this work, we ask a critical research question: Is on-policy sampling always beneficial for distilling arbitrary teacher-student pairs? We show that a simple alternative, Semi-OPD, which distills from offline rollouts generated by the initial student, can often outperform OPD in both accuracy and training efficiency. Across 17 teacher-student pairs ranging from 1.5B to 235B parameters, Semi-OPD outperforms OPD in 14 cases, with up to +13.6% accuracy and 11.4x training speedup. We further find that the choice between OPD and Semi-OPD depends on the alignment between the initial teacher and student, quantified by an output-token overlap ratio: OPD is beneficial only when t
标题：Why On-Policy Distillation Sometimes Fails: Vanishing Learning Signals
来源：arXiv
链接：https://arxiv.org/abs/2610.11247
摘要：On-policy distillation (OPD) enables effective capability transfer between language models, yet the mechanisms underlying its failures are not fully understood. Across code generation and mathematical reasoning, OPD with larger-scale teachers exhibits early loss plateaus, with an average final loss reduction of 25.1% after 200 updates, compared with 96.2% for self-RL teachers, obtained by further reinforcement learning (RL) training of the initial student. To understand this difference, we analyze OPD as an idealized continuous-time dynamical system in the small-learning-rate limit. Our training-log diagnostics associate these plateaus with an early decline in a gradient-based learning-signal proxy while substantial loss remains; these measurements do not establish why the underlying gradi
标题：RL-ARC: Calibrating Large Reasoning Models via Reasoning-guided Uncertainty
来源：arXiv
链接：https://arxiv.org/abs/2610.11352
摘要：Language models (LMs) are commonly trained with Reinforcement Learning with Verifiable Rewards (RLVR) to enhance their reasoning capabilities. However, since RLVR does not explicitly account for calibration during training, it can lead to severe calibration degradation, including overconfidence. Recent calibration-aware training methods for LMs, which incorporate objectives for uncertainty estimation into training, improve calibration but still exhibit overconfidence under distribution shift, while sacrificing reasoning performance. To this end, we propose RL-ARC, a calibration-aware training framework that jointly leverages reasoning confidence and answer confidence. Specifically, RL-ARC leverages reasoning confidence as an auxiliary signal for calibrating answer confidence, applying it a
标题：Scaling to Tens of Thousands of Test-Time Iterations with Loop-Native Attention Residuals
来源：arXiv
链接：https://arxiv.org/abs/2610.11570
摘要：In this paper, we argue that looped Transformers need their own residual connections to prevent performance degradation as the number of iterations grows. We observe that increasing loop iterations can reduce reasoning accuracy: noisy state updates overwrite correct intermediate deductions and even undo completed solutions. This leaves subsequent iterations to recover lost information from an already degraded representation: once an error arises in an earlier loop, often as a result of long-range propagation through the recurrence, later loops find it difficult to correct. In this paper, we introduce InfiLoop, a loop-native residual connection that learns which past computations to retain and how much to accept from each new update. InfiLoop combines content-based weighting with learned te
标题：Fed-GRPO: Reward-Signal-Driven Federated Group Relative Policy Optimization
来源：arXiv
链接：https://arxiv.org/abs/2610.11502
摘要：Large Language Models (LLMs) have shown strong reasoning capabilities when fine-tuned with reinforcement learning (RL), particularly through Group Relative Policy Optimization (GRPO). However, existing GRPO methods assume centralized access to training data, which may not hold in practice due to privacy or regulatory constraints. To this end, we propose Fed-GRPO, a federated GRPO training framework that addresses these privacy constraints by enabling collaborative reasoning training without sharing raw data, which leverages the reward statistics naturally produced during GRPO training as zero-cost signals to guide aggregation, local training, and communication. Fed-GRPO contains three reward-signal-driven mechanisms: (i) \emph{signal-weighted aggregation} that weights clients by their rewa
标题：From Chain-of-Thought to Loops: Non-Autoregressive Latent Reasoning via Looped Transformers
来源：arXiv
链接：https://arxiv.org/abs/2610.11472
摘要：Chain-of-thought (CoT) reasoning often improves language-model performance by giving models additional computation before answering. However, explicit CoT expresses this computation as a sequence of autoregressively generated tokens. Latent reasoning replaces these tokens with compact continuous states, but most autoregressive latent-reasoning methods retain a left-to-right dependency among latent vectors. We introduce LLoCoT: a looped latent-reasoning framework that replaces left-to-right latent generation with iterative refinement of a compact latent workspace. A shared transformer is reapplied for a small number of refinement iterations, jointly updating the latent slots based on the prompt and the evolving workspace state. Using the refined state, a probabilistic head predicts a distri
标题：SafeInferCom: Safe Inference-Time Compute via Verifier-Guided Mid-Generation Intervention for Robotic Task Planning
来源：arXiv
链接：https://arxiv.org/abs/2610.11223
摘要：Large Reasoning Language Models (LRLMs) enable multi-step reasoning for robotic task planning, but continued reasoning can overwrite valid intermediate plans or leave constraint violations unresolved, reducing planning reliability and wasting inference-time computation. We develop an inference-time monitor that exposes and verifies intermediate plans without disrupting the original decoding trajectory. Building on this monitor, we propose SafeInferCom, a formal verifier-guided framework that preserves valid intermediate plans and directs error correction during generation. Experiments across multiple LRLMs and planning domains reveal reasoning-response inconsistency and limited self-correction under one-shot inference. SafeInferCom improves planning success and accelerates error correction
标题：Environmental Feedback Modeling Matters: Rethinking Feedback Treatment in Agentic Hindsight Self-Distillation
来源：arXiv
链接：https://arxiv.org/abs/2610.11384
摘要：Reinforcement learning is commonly used to train language agents in interactive environments, but cannot be directly applied when rewards are unavailable. Recent methods use environmental feedback as privileged context for hindsight self-distillation, but our analysis suggests that simply conditioning the teacher on feedback is insufficient, motivating us to rethink how environmental feedback is used in agentic self-distillation. Given that environmental feedback contains rich supervision for modeling how the environment responds to agent actions, we introduce \textit{agentic SElf-distilLation with environmental Feedback modeling} (SELF), a framework that jointly optimizes environmental feedback modeling and hindsight self-distillation. SELF learns to predict environmental responses while
标题：Where to Adapt Matters: Layer-Selective Fine-Tuning for Capability Retention
来源：arXiv
链接：https://arxiv.org/abs/2610.11620
摘要：Parameter-efficient fine-tuning (PEFT) enables large language models (LLMs) to adapt to specialized tasks, but often at the cost of degrading general capabilities acquired during pretraining. Existing approaches primarily mitigate this trade-off through data replay or regularization, relying on additional data or explicit optimization constraints. We instead focus on a different question: where should adaptation be applied? We find that fine-tuning different Transformer layers produces different target-task gains and degrees of capability degradation, suggesting that not all layers are equally suitable for adaptation. To characterize this difference, we use layer-wise empirical Fisher information to measure target-task sensitivity. However, computing Fisher scores requires backward computa
标题：Learning the Loop, Not Just the Page: Execution-Grounded Loop Learning for Web Generation
来源：arXiv
链接：https://arxiv.org/abs/2610.11543
摘要：Functional Web generation is increasingly optimized with executable rewards, yet existing methods largely focus on the quality of the final page and leave the process of diagnosing and repairing imperfect implementations underexplored. We identify a central challenge in this setting: the Generator and Refiner produce executable artifacts with direct environment rewards, whereas the intermediate Critic influences downstream behavior without a directly executable outcome. We introduce WebLoop, an execution-grounded framework that jointly learns generation, critique, and refinement within a shared policy. WebLoop trains an execution-free Critic with complementary signals for requirement-level discriminability and downstream helpfulness, first establishing reliable diagnosis and then introduci
标题：AdaptEvo: Adaptive Agent Learning with Evolving Supervision
来源：arXiv
链接：https://arxiv.org/abs/2610.11354
摘要：Rule-governed contextual decision tasks require models to apply specified rules to case-specific context and evidence. Written rules can leave gaps in decision guidance and process evaluation, while reference judgments vary in their support from the rules and evidence. To address these challenges, we introduce AdaptEvo, a framework for learning under imperfect supervision that couples confidence-adaptive policy optimization with evolving decision knowledge and evaluation rubrics. Its Training module uses Confidence-Adaptive GRPO (CA-GRPO) to balance outcome and process rewards according to reference confidence. Its Evolution module synthesizes reusable decision knowledge from recurring failures across training cases and refines process rubrics to detect overlooked errors. To support empiri
标题：LadderEdit: Edit-Level Residual Compression for Memory-Efficient Lifelong Editing of LLMs
来源：arXiv
链接：https://arxiv.org/abs/2610.11160
摘要：Lifelong editing of LLMs requires storing thousands of edits after acquisition. A widely used family of approaches attaches one LoRA adapter per edit, which preserves behavior but grows linearly in storage. To address this challenge, we propose LadderEdit, a method that compresses each LoRA adapter after it is acquired. Each edit is first stored at low rank as a cheap sketch. We then check whether this sketch still satisfies the rewrite, generalization, and locality contract on probe prompts. Edits that pass keep the sketch; those that fail are promoted to a higher rank along a ladder until the contract is met. Because every edit retains some representation, coverage is maintained, and only hard edits consume more rank. Across ZsRE, CounterFact, and WikiBigEdit benchmarks on LLaMA-3-8B, Mi
标题：From a Prompt to Repertoires: Evolving Functional REpertoires Enable LLM Continual Learning
来源：arXiv
链接：https://arxiv.org/abs/2610.11373
摘要：Continual learning remains challenging for large language models, which must enable models to acquire new skills and knowledge without degrading existing capabilities. Existing approaches typically address this challenge by carefully designing how model parameters are updated. In contrast, prompt optimization avoids costly parameter updates while achieving competitive or even superior performance to reinforcement learning methods such as GRPO on individual knowledge-intensive and reasoning tasks. This raises a natural question: \textit{Can prompt optimization, as an efficient adaptation approach, be directly applied to continual learning?} Our analysis shows that, under sequential task adaptation, it suffers from catastrophic forgetting, while optimized prompts accumulate rules that overfi
标题：Verdict Without the Rule: Diagnosing and Auditing Regulatory Rule Sensitivity in LLM Compliance Systems
来源：arXiv
链接：https://arxiv.org/abs/2610.12313
摘要：Large language model compliance systems are deployed on the assumption that a verdict depends on the regulatory rule it is given. We test this directly across five models and 20 regulatory and platform-policy domains: delete, swap, or negate the governing rule while holding the case fixed, and check whether the verdict changes (OCS) or the model's internal representation of compliance shifts at all (ICS-delta). Neither moves much: models' verdicts are often invariant to substantial perturbations of the supplied rule, and the guard model, evaluated here under a custom-rule adaptation of its native taxonomy, is the least rule-sensitive and least accurate of the five, barely above chance (51%, versus 90-92% for general-purpose models). This reflects easy cases more than blanket neglect: on ca
标题：NanoProof: Open and Efficient Automated Theorem Proving in Lean 4
来源：arXiv
链接：https://arxiv.org/abs/2610.11605
摘要：We introduce NanoProof, to our knowledge the first factorized execution-guided theorem prover in Lean 4 whose training data, extraction tooling, training pipeline, and weights are all released, making it end-to-end reproducible using open-source resources. To this end, we build and release a dataset of structured proof trees, as well as a tool for programmatic interaction and data extraction within the Lean 4 formal verifier. To support sustainable research, we focus on compute efficiency to facilitate accessible training and evaluation. NanoProof achieves 50.8% pass@16 on MiniF2F-Test, exceeding the two closest systems of its class, HyperTree Proof Search and ABEL, at roughly 90x and 7x less compute, and using more than four orders of magnitude less compute than AlphaProof. Stronger open-
标题：Probing for Long-Horizon Deductive Reasoning Capabilities in Language Models with Prolog
来源：arXiv
链接：https://arxiv.org/abs/2610.11592
摘要：Current frontier LLMs can theoretically process long contexts with 1M tokens or more. But to what extent can they go beyond simple retrieval and perform deeper reasoning over such long contexts? We empirically investigate long-horizon reasoning capabilities of LLMs, focusing on deductive logic expressed in Prolog. We construct ProloNg, a synthetic testbed to probe Prolog Long Reasoning, which systematically varies the complexity (reasoning depth) of problems, where the hardest case has a reasoning depth of 22 and 62k context length. We study 8 reasoning models across 5 families of frontier LLMs, and find that performance degrades substantially as reasoning depth grows, with the majority of models approaching chance beyond depth 10.
标题：Workerville: Towards an Organizational Behavior Account of Agent Safety
来源：arXiv
链接：https://arxiv.org/abs/2610.11561
摘要：LLM-based agents now interact with their environments continuously, shaped by such organizational channels as user instructions, peer messages, and long-term memory. Existing safety research has examined these influences, but largely as separate agent components. How such factors jointly shape an agent's safety behavior from a unified perspective remains unmeasured. To bridge this gap, we advocate organizational behavior (OB) as a framework for studying the safety of advanced agents, reorganizing the objects of study, theoretical foundations, and experimental design around the relational structure in which agents operate. We present the first systematic formalization of counterproductive work behavior (CWB), a canonical safety-relevant subfield of OB, as Agentic Counterproductive Behavior
🔥 热点延伸 7 条 · 当日热门但非主要研究方向，仍有启发
这些条目相关性判定为 0~1 分，是有意收录的当日热点；不进入重点与研究项目分组，点击标题展开。
★0华为技术岗面试知识库与 AI 模拟面试工具开源未知 · 智能招聘 · 代码 · 原文开源项目汇集华为技术岗面经、手撕编程真题，并提供基于简历检索相似面经的 AI 模拟面试功能。
OuroWorld 是一个无需掩码的框架，可把静态 3D Gaussian Splatting 场景转化为具有丰富动态且可无限循环的 3D cinemagraph。系统先由视觉语言模型推断场景中合理的运动，再引导视频模型合成参考视频，并将其提升、补全为多视角视频。其 Inconsistency-Robust Periodic 4DGS 使用傅里叶级数形变场从结构上保证周期循环，同时通过锚定参考视角的 Grounded Drift Field 吸收跨视角不一致。该方法能够表现一般形变、物体运动和光照变化，并在 39 个重建及生成场景上取得优于基线的结果，用户研究胜率为 70.8%–99.0%。
用途与启示
为静态 3DGS 场景添加无需人工掩码的动态效果，可用于虚拟世界、数字内容和沉浸式场景制作。
周期性形变场提供了从结构上保证无缝循环的思路，可减少循环视频边界处的运动断裂。
Grounded Drift Field 展示了如何利用含噪、跨视角不一致的视频监督训练动态三维表示。
无真实动态标注的评测方案可用于衡量动态鲜活度、自然度、循环接缝一致性和场景质量。
📝 原始笔记（逐字保留）
[OuroWorld：让任意 3D 世界焕发生机，化作多样且无限循环的 3D 电影动态影像（29 ▲）](https://huggingface.co/papers/2610.12461?utm_source=digest-papers&utm_medium=email&utm_campaign=2026-10-09)
