Skip to content

AI论文速递 2026年07月25日(HuggingFace Daily Papers)

数据来源:https://huggingface.co/papers 采集时间:2026-07-25

📌 重点关注

  1. ReferTrack: Referring Then Tracking for Embodied Visual Tracking | arXiv【重点关注】 Embodied visual tracking (EVT) requires a mobile agent to continuously follow... 💡 端侧具身跟踪新范式,对移动端AI视觉感知方案有启发
  2. NVIDIA-labs OO Agents: Native Python Object-Oriented Agents | arXiv【重点关注】 Traditional agent development is split across prompt templates, tool schemas,... 💡 用Python原生对象替代prompt模板+tool schema的Agent开发范式,对Agent工程化很有参考价值
  3. OpenForgeRL: Train Harness-native Agents in Any Environment | arXiv【重点关注】 Modern AI agents rely on elaborate inference harnesses such as Claude Code, C... 💡 在CC等harness中直接训练Agent,打通推理和强化学习的闭环

📋 其他值得关注

  1. An Exam for Active Observers | arXiv — Human vision is a closed loop: gaze is continuously redirected by intermediat...
  2. Computational Humor with Multimodal LLMs: Methods, Datasets, Evaluation, and Challenges | arXiv — Multimodal humor in memes, cartoons, and comics remains difficult for AI syst...
  3. Multi-Turn On-Policy Distillation with Prefix Replay | arXiv — We study on-policy distillation (OPD) for agentic tasks, where an LLM agent i...
  4. SeededGrasp: Language-Guided Grasping in Complex Scenes with Multiple Embodiments | arXiv — Practical robotic grasping in complex scenes requires both 3D spatial reasoni...
  5. Dataset Distillation by Influence Matching | arXiv — We revisit dataset distillation from an outcome-centric perspective. Rather t...
  6. ICAE-Bench: Evaluating Coding Agents as Interactive Project Builders | arXiv — The recent emergence of vibe-coding workflows is changing what coding agents ...
  7. FinanceComplexQA: Benchmarking Agentic Reasoning on Industrial-grade Financial Documents | arXiv — Agentic Reasoning has become a transformative force in financial analysis due...