论文

本文提出 MAGE（Memory as Agent-Guided Exploration），一种将交互存储在分层状态树中的主动式执行状态管理器；实验显示，MAGE 将平均任务成功率较基线提升 7.8–20.4 个百分点，同时 token 消耗减少 55.1%。This work proposes MAGE (Memory as Agent-Guided Exploration), an active execution-state manager that stores interactions in a hierarchical state tree and shows that MAGE improves the average task success rate by 7.8--20.4 pp over baselines, while reducing token consumption by 55.1%.

结果表明，思维轨迹是推理任务的有效检索语料；将其转换为结构化、紧凑化或诊断式表征后，可释放出更强的增益。The results suggest that thinking traces are an effective retrieval corpus for reasoning tasks, and transforming them into structured, compact, or diagnostic representations unlocks even stronger gains.

本文提出 SCAR（Semantic Continuity-Aware Retrieval），一种自适应检索策略，通过权衡查询-邻居相关性与结构连续性惩罚来选择性扩展相邻分块，由此得到近似尺度不变的决策规则，无需重新校准即可跨 embedding 模型迁移。SCAR (Semantic Continuity-Aware Retrieval), an adaptive retrieval policy that selectively expands neighboring chunks by weighing query-neighbor relevance against a structural continuity penalty, is proposed, yielding an approximately scale-invariant decision rule that transfers across embedding models without recalibration.

本文提出 PathRouter，一种面向 agentic GraphRAG 的路径感知训练框架，沿答案正确性与证据路径重叠度联合评估每条轨迹，归纳出四类轨迹并采用差异化 GRPO 优势缩放，抑制捷径式强化同时保留证据寻求行为。This work presents PathRouter, a path-aware training framework for agentic GraphRAG that jointly evaluates each trajectory along answer correctness and evidence-path overlap, yielding four trajectory categories with differentiated GRPO advantage scaling that suppresses shortcut reinforcement while preserving evidence-seeking behavior.

本文提出 User as Code（UaC）范式，将 Agent 对用户的建模视为一个活的软件项目：类型化 Python 对象承载用户状态，普通 Python 函数编码其治理规则，从而在解释器可运行的同一媒介内完成对用户的表示与推理。This work introduces User as Code (UaC), a paradigm in which an agent's model of a user is a living software project: typed Python objects hold the user's state and ordinary Python functions encode the rules that govern it, so representing and reasoning about the user happen in one medium an interpreter can run.

本文提出 Diverge，一种即插即用的 agentic RAG 框架，通过迭代式、反思引导的多视角探索以及多样性感知检索支持来改善多样性—质量权衡，并引入用于刻画开放域问答中多样性—质量权衡的评估指标。Diverge is proposed, a plug-and-play agentic RAG framework that improves the diversity--quality trade-off through iterative, reflection-guided exploration of diverse viewpoints and diversity-aware retrieval support, and introduces evaluation metrics for characterizing the diversity-quality trade-off in open-ended question answering.

本文命名并研究这两条研究脉络之间的模式：其递归单元是配备文件系统工具、代码执行与规划的完整 Agent harness，而非无工具的模型调用，并给出针对长上下文推理的受控评估。This work names and studies the pattern between these two lines of work, where the recursive unit is a full agent harness with filesystem tools, code execution, and planning rather than a model call with no tools, and provides a controlled evaluation on long-context reasoning.

作者的分析揭示了基于扩展设计的基本局限：扁平化层级会导致 PE-Online 中递归查询延迟过高，并在两种扩展策略下产生结构变更时不可扩展的写放大；与之相对，TrieHI 将目录拓扑保留为原生前缀树，通过树遍历实现高效递归检索，借助拓扑节点操作降低维护成本。The authors' analysis exposes the fundamental limitations of expansion-based designs: flattening the hierarchy incurs high recursive-query latency in PE-Online and unscalable write amplification during structural changes in both expansion strategies, and in contrast, TrieHI keeps the directory topology as a native prefix tree, enabling efficient recursive retrieval through tree traversal and reducing maintenance cost through topological node manipulation.

本文提出名为 InSemRAG 的 RAG 框架，通过迭代检索—校验机制及两个支撑模块——意图感知检索器（IAR）与语义保持切分（SPC）——应对上述挑战。This work proposes a RAG framework, termed InSemRAG, that addresses these challenges via an iterative retrieve-and-check mechanism with two supporting modules, an intention-aware retriever (IAR) and semantics-preserving chunking (SPC).

MRAgent，一种将联想记忆图与主动重构机制相结合的框架，将 LLM 推理直接融入记忆访问，确保记忆检索能动态适配推理上下文，同时避免无约束扩展引发的组合爆炸。MRAgent, a framework that combines an associative memory graph with an active reconstruction mechanism that integrates LLM reasoning directly into memory access, ensuring that memory retrieval is dynamically adapted to the reasoning context while avoiding combinatorial explosion caused by unconstrained expansion.

本文提出 Vision-aligned Latent Reasoning（VaLR），一种简洁而有效的推理框架，在每个 Chain of Thought 推理步骤之前动态生成视觉对齐的 latent token，引导模型在 latent space 中基于感知线索进行推理。Vision-aligned Latent Reasoning (VaLR) is introduced, a simple, yet effective reasoning framework that dynamically generates vision-aligned latent tokens before each Chain of Thought reasoning step, guiding the model to reason based on perceptual cues in the latent space.

本文提出 ContextRL，一种上下文感知的强化学习方法，通过间接辅助目标提升长周期推理与多模态性能，并与将相同对比上下文复用作标准 query–context–answer 样本的数据增强基线进行对比。This work proposes ContextRL, a context-aware reinforcement learning (RL) method that improves long-horizon reasoning and multimodal performance through an indirect auxiliary objective, and compares against data-augmentation baselines that repurpose the same contrastive contexts as standard query--context--answer examples.

本文首次系统分析了 LLM 如何通过强化学习（RL）学习针对不同 retriever 调整 query 表述策略，并揭示 RL 能有效教会 LLM 根据特定 retriever 特性定制 query。This work presents the first systematic analysis of how LLMs can learn to adapt their query formulation strategies for different retrievers via reinforcement learning (RL), and reveals that RL effectively teaches an LLM to tailor its queries to specific retriever characteristics.

SwiftCache 是一个协同推理系统，使异构模型可在同一服务器内共享未充分利用的 GPU 内存与 NVLink 带宽，支持跨模型通过 NVLink 共享 KV cache，避免使用慢速 PCIe 传输。SwiftCache is a collaborative inference system that enables heterogeneous models to share underutilized GPU memory and NVLink bandwidth within a server, allowing cross-model KV cache sharing over NVLink and avoiding slow PCIe transfers.

一个两级框架将手动 harness 工程转变为自动化 harness 工程，并更进一步——将"自动化本身的设计"也自动化。A two-level framework shifts manual harness engineering into automated harness engineering, and takes one step further --automating the design of the automation itself.

综合结果表明，强大的 coding agent 通过使用工具、反馈和工作区状态来构建目标语言的可工作模型，从而适应不熟悉的语言。Together, these results show that strong coding agents adapt to unfamiliar languages by using tools, feedback, and workspace state to build a working model of the target language.

提出 Agentic Harness Engineering（AHE），一个通过三个相互匹配的 observability 支柱应对 harness 工程挑战的闭环，将每一次编辑转化为可证伪的契约，使 harness 演进能够自主进行而不退化为试错。Agentic Harness Engineering (AHE) is introduced, a closed loop that addresses harness engineering challenges through three matched observability pillars that turn every edit into a falsifiable contract, so harness evolution proceeds autonomously without collapsing into trial-and-error.

提出 OScaR（Omni-Scaled Canalized Rotation），一个面向 X-LLMs 的精确且轻量的 KV cache 压缩框架，作为一个鲁棒、低复杂度、通用的框架确立了新的 Pareto 前沿。OScaR (Omni-Scaled Canalized Rotation), an accurate and lightweight KV cache compression framework for X-LLMs, is proposed, establishing it as a robust, low-complexity, and universal framework that defines a new Pareto front.

本文提出 Natural-Language Agent Harnesses，即可编辑的、描述运行级 harness 策略的文档，以及 Intelligent Harness Runtime（IHR），一个将上述文档解释为 agent 调用、交接、状态更新、验证门控与 artifact 契约的共享运行时。This paper introduces Natural-Language Agent Harnesses, editable documents that describe run-level harness policy, and Intelligent Harness Runtime (IHR), a shared runtime that interprets these documents into agent calls, handoffs, state updates, validation gates, and artifact contracts.

提出 agent harness 的操作性定义，配套统一词汇，可用于指导工程实践以及对 agentic 系统的科学比较。An operational definition of agent harness is proposed, with a shared vocabulary, able to guide engineering practice and the scientific comparison of agentic systems.

WAIT (Waiting for Accumulated Inference Threshold) 是一种基于阈值的准入规则，适用于已知输出长度；Nested WAIT 通过调控请求在 decode 阶段各分段间的推进方式，将该规则扩展到未知输出长度场景。WAIT (Waiting for Accumulated Inference Threshold), a threshold-based admission rule for known output lengths, and Nested WAIT, which extends the rule to unknown output lengths by regulating how requests advance across decode-stage segments are designed.

本文旨在应对常见挑战并简化 RAG 与既有企业基础设施的集成，提出一套面向可扩展本地化部署企业 RAG 方案的综合性 AI 工程蓝图。This paper aims to address the gap in comprehensive on-premises RAG implementation frameworks by presenting a comprehensive AI engineering blueprint for scalable on-premises enterprise RAG solutions to address common challenges and streamline the integration of RAG into existing enterprise infrastructure.

本文提出 AIConfigurator，一个统一的性能建模系统，能够在不依赖 GPU profiling 的前提下进行快速、与框架无关的推理配置搜索；并提供一个抽象层，自动为目标后端解析最优启动参数，无缝集成到生产级编排系统中。AIConfigurator is presented, a unified performance-modeling system that enables rapid, framework-agnostic inference configuration search without requiring GPU-based profiling, and an abstraction layer that automatically resolves optimal launch parameters for the target backend, seamlessly integrating into production-grade orchestration systems.

GLM 将推理分解为分类、推理、动作生成与图检索等专用 Agent，支持分支化与选择性上下文共享，在保持推理质量的同时降低 prompt 长度与推理迭代次数，从而提升准确率并减少整体 token 消耗。GLM decomposes reasoning into specialized agents for classification, reasoning, action generation, and graph retrieval, enabling branching and selective context sharing to reduce prompt length and reasoning iterations while preserving reasoning quality, thereby improving accuracy and reducing overall token consumption.

研究发现 RLMs 能够成功处理超出模型上下文窗口长达两个数量级的输入，即便在较短 prompt 下，其质量也显著优于原生前沿 LLM 以及常见的长上下文与编程脚手架。It is found that RLMs can successfully process inputs up to two orders of magnitude beyond model context windows and, even for shorter prompts, dramatically outperform the quality of vanilla frontier LLMs and common long-context and coding scaffolds.

本文提出一个多 Agent 框架，由主 LLM 协调 grounding agent 定位与问题相关的片段，并协调 vision agent 提取目标文本观测，显著优于强非 Agent 基线，且强化学习进一步增强了已训练 Agent 的推理与规划能力。A multi-agent framework in which a master LLM coordinates a grounding agent to localize question-relevant segments and a vision agent to extract targeted textual observations is proposed, which significantly outperforms strong non-agent baselines and shows reinforcement learning further strengthens reasoning and planning for the trained agent.

本文提出 ReMemR1，将 memory retrieval 机制融入 memory update 过程，使 Agent 能够选择性回溯历史记忆以进行非线性推理，并设计多级奖励机制，将最终答案奖励与密集的 step 级信号相结合，引导对 memory 的有效使用。This work presents ReMemR1, which integrates the mechanism of memory retrieval into the memory update process, enabling the agent to selectively callback historical memories for non-linear reasoning, and proposes a multi-level reward design, which combines final-answer rewards with dense, step-level signals that guide effective memory use.

本研究建立了一套实用的 LongPT 方案，为推进长上下文 vision-language 模型奠定了经验基础，并提出 MMProLong，无需任务专属监督即可泛化至基于网页的多模态 needle 检索、长上下文图文压缩以及长视频理解等任务。This study establishes a practical LongPT recipe and an empirical foundation for advancing long-context vision-language models, and introduces MMProLong, which generalizes to webpage-based multimodal needle retrieval, long-context vision-text compression, and long-video understanding without task-specific supervision.

本文提出 LLaDA-V，一种完全基于扩散范式的多模态大语言模型 (MLLM)，将视觉指令微调与 masked diffusion 模型相结合，脱离了当前多模态方法中主流的自回归范式。LLaDA-V is introduced, a purely diffusion-based Multimodal Large Language Model (MLLM) that integrates visual instruction tuning with masked diffusion models, representing a departure from the autoregressive paradigms dominant in current multimodal approaches.

Decentralized Language Models (DeLM) 是一种 MAS 框架，通过并行 Agent、共享已验证上下文与任务队列实现去中心化协作，在软件工程 test-time scaling 与长上下文推理两方面均取得提升。Decentralized Language Models (DeLM), a MAS framework that decentralizes coordination through parallel agents, a shared verified context, and a task queue, improves both software-engineering test-time scaling and long-context reasoning.

提出一种自我演化的法律 Agent 框架，将 Model、Harness、Agent 角色、法律 Knowledge、确定性 Tools 与程序性 Skills 拆分为可审计的层面，用于来源可追溯性、日期与数字的事实校验、交付物合规性以及问题闭环。A self-evolving legal-agent framework that factors Model, Harness, Agent roles, legal Knowledge, deterministic Tools, and procedural Skills into auditable surfaces for source traceability, date and number grounding, deliverable compliance, and issue closure is introduced.

引入 D（Domain-Collection-Document），一种面向领域的设计，用于在不修改底层语言模型的前提下组织 RAG 系统中的知识并控制查询处理。D (Domain-Collection-Document), a domain-oriented design to structure knowledge and control query processing in RAG systems without modifying the underlying language model, is introduced.

提出 Tail-Aware Adaptive-k（TAA-k），一个无需训练、通过局部验证策略将 EVT 落地的框架，检索质量接近 oracle，相较全局 EVT 方法获得数量级的效率提升，并在不同 embedding 模型和压缩维度下保持稳健。Tail-Aware Adaptive-k (TAA-k), a training-free framework that operationalizes EVT through a localized validation strategy, is proposed, which achieves near-oracle retrieval quality with orders-of-magnitude efficiency gains over global EVT methods, while maintaining robustness across embedding models and compression dimensions.

本文提出后确定性分布式系统（PDDS）作为一种研究与工程模型，用于协调确定性代码、随机模型与自主 Agent 共存的异构环境，并证明经典分布式计算模型是该参与者通用模型的零模糊特例。This paper introduces Post-Deterministic Distributed Systems (PDDS) as a research and engineering model for coordinating heterogeneous environments where deterministic code, stochastic models, and autonomous agents coexist, and shows that classical distributed computing models form a zero-ambiguity special case of this participant-general model.

提出 Stratum，一种统一系统基础设施，在 agentic pipeline 搜索期间将 pipeline 执行与规划和推理解耦，无缝集成现有 Python 库，将批量 pipeline 编译为优化的执行图，并通过基于 Rust 的新型 runtime 等异构后端高效执行。Stratum is proposed, a unified system infrastructure that decouples pipeline execution from planning and reasoning during agentic pipeline search, and integrates seamlessly with existing Python libraries, compiles batches of pipelines into optimized execution graphs, and efficiently executes them across heterogeneous backends, including a novel Rust-based runtime.

HERA 是一个分层框架，联合演化多 Agent 编排与角色特定的 Agent prompt，在多 Agent RAG 中实现针对性的、角色条件化的改进，并提供角色感知的 prompt 演化。HERA is a hierarchical framework that jointly evolves multi-agent orchestration and role-specific agent prompts, enabling targeted, role-conditioned improvements in multi-agent Retrieval-Augmented Generation and Role-Aware Prompt Evolution.