👍 185
08/25 20:00
A common strategy for scaling world models is to train on more crawled video with more compute. We argue that this strategy is inefficient: scaling world models also requires a recursive data engine that offers grounded reward signals. The success of code agents illustrates why this matters. As code
中文介绍 论文提出了一种可验证的轨迹数据引擎,以提高世界模型的扩展能力。通过实现一个递归的数据引擎,提供有基础的奖励信号,从而解决单一依赖视频爬虫和计算能力的低效问题。具体来说,以代码代理的成功为例,论证了这种方法在提升模型性能中的重要性。意义:对世界模型和自主代理的发展具有重要的启示。
👍 106
08/26 20:00
Multimodal large language models (MLLMs) can interpret a street view, but urban agency depends on whether such local evidence remains useful after the agent starts to move. In this paper, we investigate how far current MLLM agents can turn local urban perception into reliable action in a complicated
中文介绍 本研究探讨了多模态大型语言模型在复杂城市环境中的行动能力,重点分析了局部城市感知如何转化为可靠的行动。通过实地考察不同场景,分析了当前模型在适应移动后的局部证据利用情况的局限性,为提升城市智能代理的有效性提供了可能的方向。意义:推动城市智能体在动态环境中的应用研究进展。
👍 82
08/27 20:00
Loop Engineering is emerging as a practice for organizing development work around coding agents. Instead of writing each prompt by hand, practitioners design loops that monitor progress, assign work, run checks, and decide what the agent should do next. Even with a capable coding agent, a loop may t
中文介绍 论文介绍了Loop Engineering作为一种新的开发实践,旨在围绕编码代理组织工作。与手动编写每个提示不同,该方法通过设计循环来监控进度并制定下一步行动,从而提高开发效率。尽管能力较强的编码代理仍面临挑战,该方法为编码工作提供了创新的管理方式。意义:推动智能体协作和自主开发工作的进展。
👍 72
08/26 20:00
Recent prominent post-training methods, such as Reinforcement Learning (RL) and On-Policy Self-Distillation (OPSD), have driven rapid progress in mathematical reasoning for large language models, yet their reliance on ground-truth labels precludes test-time training (TTT). Replacing ground truth wit
中文介绍 研究提出了一种新的测试时策略优化(TTPO)方法,旨在克服传统后训练方法对真实标签的依赖,进而实现测试时训练(TTT)。通过模拟环境中生成伪标签,该方法能够提升模型在实际应用中的适应性和灵活性,为大规模语言模型在推理和决策中的应用提供了新的思路。意义:对推理和自适应学习领域的研究具有深远影响。
👍 60
08/18 20:00
Equipping Large Language Models (LLMs) with multi-turn tool-calling capabilities is essential for building autonomous agents. However, progress is fundamentally limited by the reliance on full-length trajectory imitation. For tasks involving multiple order-independent sub-goals, the optimal solution
中文介绍 本研究针对大型语言模型的多轮工具调用能力向自主代理的演变进行探索,提出了一种新的自我蒸馏方法。通过提升对全长轨迹模仿的理解,旨在打破限制代理发展的现有限制,为处理无序子目标任务提供更优解决方案。意义:为智能体的自主学习和适应能力提供新途径。
👍 59
08/26 20:00
Scaling robot data is crucial for building generalist Vision-Language-Action (VLA) models, yet robot trajectories are harder to scale than web-scale image-text data because embodied collection is costly and sparsely covers the physical world. This makes representation quality a central bottleneck: u
中文介绍 论文探讨了在机器人视觉-语言-行动(VLA)模型训练中,数据扩展面临的挑战及其对表现的影响。特别强调在实际环境中获得轨迹数据的高成本,使得表现的质量成为模型扩展的核心瓶颈。本研究为提升机器人在复杂环境中的适应能力提出了新的思路和方法。意义:推动通用VLA模型的研究和应用。
👍 51
08/27 20:00
Generative models can turn natural-language prompts into images, text, code, and other content, lowering the cost of producing drafts and components. Their practical impact increasingly depends on whether those pieces can become complete, dependable deliverables. This survey examines agentic artifac
中文介绍 调查了生成模型在自然语言提示转化为可交付成果的有效性,强调其对创意输出的实践影响。通过对系统、评估和原理的综合考察,论述了如何确保代理生成内容的可靠性与完整性,为智能体生成艺术品的实际应用奠定基础。意义:为生成艺术与设计领域的智能化提供新视角与机会。
👍 47
08/21 20:00
AI-powered digital avatar streamers must answer product questions, engage viewers, and execute marketing strategies in real time, demanding low latency, frequent strategy updates, and accurate yet effective responses. Evolvable Harnesses, whose Skills, Hooks, prompts, and tools can be updated indepe
中文介绍 研究提出了一种AI驱动的数字化头像代理,旨在实时处理产品咨询和观众互动。通过可进化的控制框架,该代理能够保持低延迟与精准响应,提升营销策略的灵活性与效果,为流媒体和互动应用中的智能代理提供新思路。意义:推动智能代理在实时互动中的应用与发展。
👍 45
08/24 20:00
Modern video games combine first-person perception, rapid visual changes, persistent world state, and heterogeneous native controls. Existing game agents map visual and task context directly to actions but lack explicit world dynamics modeling, whereas interactive game world models predict visual fu
中文介绍 论文提出一种新的世界动作模型(WAM),专为现代视频游戏设计。这种模型结合第一人称感知和迅速变化的视觉状态,能够动态预测游戏世界中的视觉变化,弥补已有游戏代理在建模世界动态方面的不足。意义:为视频游戏智能代理的开发提供先进的方法论。
👍 43
08/26 20:00
Physical understanding and reasoning depend on forming compact and generalizable representations of the world. While modern vision-language models can recognize and explain diverse physical events, they often lack explicit representations of the underlying mechanisms-such as object states, physical
中文介绍 研究探讨了物理推理中对可执行世界表征的自主发现能力,提出了一种新的方法来提升模型的物理理解和推理能力。尽管现有的视觉-语言模型能够识别和解释物理事件,但往往缺乏对 underlying 机制的清晰表征。意义:助力物理推理与智能体开发的协同进步。
👍 35
08/26 20:00
Self-evolving language models have recently emerged as a promising path toward superintelligence, with the advantage of reducing the cost of human supervision. While considerable progress has been made in verifiable domains, self-evolution in unverifiable domains remains substantially less explored.
中文介绍 研究探索了零数据自我进化的语言模型,强调在没有人类监督情况下实现智能体的演变能力。尽管在可验证领域已取得进展,但在不可验证领域内,自我进化的研究仍然较少,为未来的发展提供了新方向。意义:为自我学习与演化智能体的研究拓展边界。
👍 30
08/26 20:00
Long-horizon agent runs generate experience that can improve both the current run and future work. Most self-improvement methods process this experience only after execution ends, so they cannot redirect the active run or immediately apply and validate lessons learned from it. We argue that self-imp
中文介绍 论文探讨了长期代理在执行过程中的自我提升能力,提出通过即时反馈来改进当前运行和未来任务的方法。大多数传统自我提升方法只能在执行结束后处理经验,这限制了其灵活性和实时性。因此,该研究强调了实时自我改进的重要性。意义:推动智能体的自我学习和实时反馈机制的创新。
👍 27
08/26 20:00
Language model pretraining has become almost synonymous with prohibitive cost, placing it out of reach for much of the academic and open-source communities. Although strong open-source efforts already exist, including open-weight models and open-source training recipes, a cost-efficient, hardware-ac
中文介绍 本研究提出了Puro-2B,旨在降低语言模型预训练的成本,从而使更多研究者能够访问该技术。通过优化硬件的使用,该模型能在有限资源下进行有效训练,为开放源和学术社区提供了实践基础。这对降低AI训练成本具有重要的意义。意义:为开放源技术和资源公平化提供新的可能。
👍 24
08/23 20:00
Vision-Language-Action (VLA) models can turn multimodal context into robot actions, but their action decoders are still trained largely by behavior cloning. This supervises which motor command was demonstrated while leaving implicit the local objective served by the behavior under the instruction. F
中文介绍 论文探讨了Vision-Language-Action(VLA)模型中行为意图的提炼,旨在增强机器人行为解码的效果。通过分析当前模型主要依靠行为克隆的局限性,提出了一种更具目标导向的监督方式,推动机器人行动解码的智能化进程。意义:为多模态学习和智能体行为建模提供新思路。
👍 24
08/27 20:00
Long-horizon agentic tasks require large language models (LLMs) to iteratively retrieve, integrate, and maintain dispersed information across multi-turn interactions, but preserving all interaction histories leads to a continuously growing working context. Recent proactive context management methods
中文介绍 研究重点分析了长时间任务中的主动上下文管理对模型性能的影响,指出当前方法在处理多轮交互时面对上下文增长瓶颈。提出了一种细粒度的强化学习方法,以提高长时间任务中的上下文管理能力,提升智能体的响应效率和准确性。意义:对长时间任务智能体中的信息管理提供重要参考。
👍 24
08/26 20:00
Agent skills package specialized knowledge and workflows into reusable resources that extend AI agent capabilities. Recent work automatically discovers such skills from agent experience, which enables agents to progressively adapt through interaction. However, the insights that guide skill developme
中文介绍 本研究提出了WikiSkill,通过将代理经验编译为持久知识,实现技能的进化。该方法赋予代理具有适应性的能力,使其能够在互动中不断提升技能。然而,对技能发展的指导见解仍待进一步深入。意义:推动智能体在复杂任务中的技能自我提升与演化。
👍 19
08/25 20:00
Zero-shot cross-task generalization, where a policy must execute manipulation tasks never seen during training, remains a central challenge in robot learning. In large language models, a novel task can be performed simply by specifying it in the context, without any parameter update. This form of in
中文介绍 论文探讨了跨任务的零-shot泛化问题,特别是在机器人学习中未见过的操作任务的执行能力。提出了一种新颖的上下文内世界-行动建模方法,能够在未进行参数更新的情况下,直接通过上下文描述执行新任务,为机器人学习提供了新方法。意义:为机器人任务泛化和灵活应用开辟重要路径。
👍 15
08/27 20:00
Spatio-temporal video grounding (STVG) requires models to identify when a referred event occurs and localize the target entity throughout that interval. Existing multimodal large language models typically serialize dense localization trajectories autoregressively, causing decoding latency to grow wi
中文介绍 研究针对时空视频定位(STVG)中的有效生成策略进行探讨,重点分析现有多模态大型语言模型在识别事件发生时间和目标实体定位中的局限性。通过重塑生成过程,旨在提升视频定位的效率,从而为视频分析和检索提供更好的解决方案。意义:推动视频智能技术的进步与应用。
👍 15
08/27 20:00
Factual question answering (QA) typically assumes a single canonical answer, obscuring whether large language models (LLMs) retain divergent accounts of long-tail facts. To address this gap, we introduce ElephantBench, a closed-book knowledge probe comprising 1,094 questions generated through an aud
中文介绍 论文提出了ElephantBench,这是一个新的知识探测框架,旨在评估大语言模型在处理长尾知识时的能力。通过946个基于真实情况生成的问题,研究揭示了当前模型在事实问答中的不足,为未来模型的改进提供了依据。意义:为长尾知识的处理与模型评估提供了新视角。
👍 14
08/25 20:00
Evaluation is shifting from static QA toward agentic settings where models act through external tools. We identify a critical yet underexplored capability within this space - dexterous visual tool use: fine-grained, closed-loop parameterized visual action in which models infer tool parameters from v
中文介绍 研究聚焦于多模态代理在灵巧视觉工具使用中的性能评估,确定了该领域一个关键能力,即在闭环参数化视觉动作中推理工具参数。该研究为在智能体与外部工具交互的环境中性能评估提供了新的基准,具有实际应用价值。意义:推动智能代理在复杂操作中的灵活应用与发展。