👍 186
09/23 08:00
Object permanence and solidity are hallmarks of human cognitive priors. Recent studies show that video generation models, a paradigmatic class of current world models, have begun to show emerged reasoning abilities, making them ideal candidates for building human-like physical intelligence. Do video
中文介绍 针对视频生成模型是否具备类人认知先验的问题,本文研究在 world models 中训练客体永久性与固体性。通过设计特定训练任务与数据,探索视频生成模型能否习得物体在遮挡后仍持续存在的推理能力,为构建类人物理智能提供依据。
👍 82
09/22 08:00
Long-term conversational memory in multi-party settings requires more than retrieving relevant content from long-term conversations: it must distinguish who said what, whom each statement concerns, how individuals perceive one another, what information is shared by the group, and how states change o
中文介绍 多轮多方对话的长期记忆需区分谁说了什么、涉及谁、群体共享信息及状态演变。本文提出 SpeakerMem-R1,采用以说话人为中心的双轨记忆机制,分别建模个体与群体层面信息,提升多方对话中长期记忆的检索与推理能力。
👍 55
09/24 08:00
While Large Language Models (LLMs) rely on highly non-linear components, in this work we demonstrate that they exhibit fundamental linearity: when inputs from distinct text streams are linearly combined, the model outputs a superposition of the individual next-token distributions. We term this the S
中文介绍 本文发现 LLM 虽含高度非线性组件,却表现出基础线性:将不同文本流的输入线性组合后,模型输出为各自下一 token 分布的叠加,称为线性叠加现象。该发现揭示了 Transformer 内部表征的线性结构,为理解与操控 LLM 提供新视角。
👍 45
09/19 08:00
Spatial reasoning is essential for vision-language models (VLMs) to understand and act in the physical world. Reasoning in dynamic environments requires VLMs to perceive local state transitions caused by object motion and viewpoint changes and integrate them over long trajectories to maintain an upd
中文介绍 针对 VLM 在动态环境中空间推理不足的问题,本文提出 Spatial-Interactor,通过与可观测物理世界交互来学习空间推理。模型感知物体运动与视角变化引起的局部状态转移,并在长轨迹上整合以维持更新表征,提升具身空间理解能力。
👍 40
09/21 08:00
Evaluating world models requires assessing both the quality of the worlds they generate and their consistency and responsiveness under exploration, interaction, and modification. We introduce HappyWorld-Bench, a comprehensive benchmark that evaluates whether generated worlds remain reliable as agent
中文介绍 评估 world models 需兼顾生成质量与探索、交互、修改下的一致性和响应性。本文提出 HappyWorld-Bench 综合基准,检验生成世界在智能体交互中是否保持可靠,为 world models 的系统评估提供统一标准与测试维度。
👍 38
09/23 08:00
Advances in generative models have improved video fidelity, enabling long-horizon generation, interactive world modeling, and evolving visual environments. Autoregressive (AR) video generation extends visual sequences through causal rollouts. However, a fundamental bottleneck emerges: as the generat
中文介绍 自回归视频生成通过因果 rollout 扩展视觉序列,但随生成推进会出现根本性瓶颈。本文提出以历史帧为记忆约束未来生成,缓解长时程自回归视频中的漂移与一致性退化,提升交互式世界建模的长期生成质量。
👍 34
09/23 08:00
Agentic memory systems reuse past experience to improve future performance, yet most existing designs curate memory at write time: once a task is completed, its trajectory is distilled into a fixed artifact, such as a reflection, workflow, skill, or reasoning strategy, that is later retrieved by sim
中文介绍 现有智能体记忆多在写入时固化经验为反思、工作流或技能等固定产物,检索时难以适配新任务。本文提出 Just-in-Time Memory,在推理时学习按任务动态整编记忆,使 LLM Agent 能复用经验并自适应检索,提升未来任务表现。
👍 20
09/20 08:00
Humans can effortlessly localize the direction of a sound source and integrate it with visual cues for reasoning, yet this remains challenging for embodied agents. In particular, it is still unclear how to effectively evaluate and model spatial audio understanding in embodied settings. To address th
中文介绍 人类能轻松定位声源方向并与视觉线索融合推理,具身智能体却面临困难。本文提出 OmniEcho,面向具身场景建模空间音频理解,设计评估与建模方法,使智能体结合空间音频与视觉信息进行推理与决策。
👍 16
09/22 08:00
Reinforcement learning has become a central component of large language model (LLM) post-training, yet token-level credit lacks a generally accepted mathematical definition, leaving its relationship to commonly used training signals unclear. We formulate three regularity conditions, namely Completen
中文介绍 针对 LLM 后训练中 token 级 credit 缺乏统一定义的问题,本文提出 PACT 框架,从信用分配走向 critic 对齐。通过形式化完整性等三条正则条件,厘清 token 级信用与常用训练信号的关系,为 RL 后训练提供更严谨的理论基础。
👍 14
09/22 08:00
The rapid capability gains of frontier language models are widely attributed to improved reasoning abilities, yet this cannot be verified as raw CoT traces in closed-source systems are hidden. By registering a simple custom tool through a standard API feature, we induce frontier models to externaliz
中文介绍 闭源前沿模型的原始 CoT 被隐藏,其推理能力提升难以验证。本文通过标准 API 注册自定义工具,诱导前沿模型外化隐藏的思维链,从而提取并刻画其隐式推理过程,为评估闭源系统推理能力提供新手段。
👍 13
09/23 08:00
Recent advances in large language models (LLMs) have enabled agents to tackle long-horizon tasks across diverse environments. To further improve agent performance, existing language world models typically predict environment observations, yet reconstructing high-entropy, execution-dependent tool res
中文介绍 现有语言 world models 多预测环境观测,却需重建高熵、依赖执行细节的工具结果,成本高且噪声大。本文提出 Agent-Editing World Model,重构面向 LLM Agent 的世界建模方式,减少对观测重建的依赖,提升长时程任务中的智能体性能。
👍 10
09/20 08:00
Robotic bin packing requires long-horizon sequential decision-making, as each object placement affects the available space for subsequent packing. Existing methods primarily rely on hand-crafted geometric heuristics that optimize predefined objectives or reinforcement learning policies learned throu
中文介绍 机器人装箱需长时程顺序决策,每步放置影响后续空间。现有方法多依赖手工几何启发式或强化学习策略。本文提出 PackLab 框架,支持 MLLM 在机器人装箱任务中的开发、训练与评估,推动数据驱动装箱决策研究。
👍 10
09/23 08:00
Vision-Language-Action models provide a strong foundation for general-purpose robot control, yet a vast majority of policies do not preserve and leverage episode-level information beyond the current observation. This limitation is consequential in history-dependent manipulation tasks that depend on
中文介绍 多数 Vision-Language-Action 策略仅依赖当前观测,无法保留和利用回合级历史信息,在依赖历史的操作任务中受限。本文提出 MemBodied,为 VLA 模型引入循环联想记忆,实现跨观测的记忆保持与调用,提升历史依赖操作任务表现。
👍 9
09/24 08:00
The rapid progression of large language models is extending AI from passive content generation into the active workflows of engineering and scientific discovery. This shift raises a compelling question: can AI be both the object of development and an active participant in building next-generation AI
中文介绍 面向真实移动端规划智能体,本文提出 Qwen-Planner-Agent,一个闭环 AI-for-AI 框架,让 AI 既作为开发对象又参与构建下一代 AI。该框架通过闭环迭代优化规划器,提升移动智能体在真实场景中的规划与执行能力。
👍 9
09/23 08:00
We present Hunyuan-A13B, an open-source large language model based on a Mixture-of-Experts architecture. It contains 80 billion total parameters but activates only 13 billion during inference, balancing model capability, computational efficiency, and deployment cost. The model is pretrained on a rig
中文介绍 本文发布开源大模型 Hunyuan-A13B,采用 Mixture-of-Experts 架构,总参数 800 亿、推理仅激活 130 亿,兼顾能力、计算效率与部署成本。模型基于大规模数据预训练,并配套后训练流程,以较小激活量实现有竞争力的性能。
👍 8
09/24 08:00
Deep search requires LLM agents to decompose complex queries, search for evidence, and synthesize grounded answers, yet existing ReAct-style agents suffer from two limitations: role coupling, where one policy must handle planning, evidence use, and synthesis; and context accumulation, where growing
中文介绍 现有 ReAct 式深度搜索智能体存在角色耦合与上下文累积问题,单一策略需兼顾规划、证据使用与综合。本文提出 IterSynth,通过角色解耦的迭代合成方法,分离各角色并使用合成经验,提升深度搜索任务的分解、检索与答案生成能力。
👍 8
09/23 08:00
Language-model agents increasingly face long-horizon tasks with evolving state, interdependent decisions, and delayed outcomes. Scaling their training requires diverse agentic environments, dependable outcome signals, and low extension cost. Existing generation pipelines commonly construct an enviro
中文介绍 训练长时程语言模型智能体需多样环境、可靠结果信号与低扩展成本。本文提出 Verifiable Hidden Dynamics Play,从已解机制生成智能体 RL 环境,提供可验证结果信号,降低环境构建成本,扩展智能体强化学习训练规模。
👍 8
09/23 08:00
AI research agents need reliable knowledge of how their experiments change outcomes. We introduce WhatWorkedBench to measure experimental understanding, the accuracy of predictions about component changes after budgeted experimentation. Agents inspect code, select measurements, and submit a response
中文介绍 AI 科研智能体需理解实验改动如何影响结果。本文提出 WhatWorkedBench,衡量实验理解能力,即有限预算实验后对组件变更结果的预测准确度。智能体需检查代码、选择测量并作答,为评估科研 Agent 的实验理解提供基准。
👍 7
09/24 08:00
Rufus-Air is an open and reproducible post-training recipe on GLM-4.5-Air-Base (106B-A12B), organized as a serial pipeline of eight stages: SFT, Reasoning RL, Coding RL, Instruction-Following RL, General Agent, Coding Agent, Search Agent, and RLHF. We document the data, reward design, infrastructure
中文介绍 本文提出 Rufus-Air,一个基于 GLM-4.5-Air-Base(106B-A12B)的开源可复现后训练配方,包含 SFT、推理 RL、代码 RL、指令遵循 RL、通用 Agent、代码 Agent、搜索 Agent 与 RLHF 八阶段,并公开数据、奖励设计及基础设施细节。
👍 6
09/24 08:00
In this paper, we propose RGBD20K, a novel dataset for facilitating the development of more robust and general RGB-D semantic segmentation by encompassing abundant categories and high-quality annotations. RGBD20K possesses several attractive properties: (1) Expanded Semantic Space. In particular, it
中文介绍 本文提出 RGBD20K 大规模 RGB-D 语义分割数据集,涵盖丰富类别与高质量标注,扩展语义空间,旨在推动更鲁棒、通用的 RGB-D 语义分割研究,为多模态深度感知分割提供新的基准与训练资源。