Skip to content

AI论文速递 2026年05月01日（HuggingFace Daily Papers）

数据来源：https://huggingface.co/papers 采集时间：2026-05-01

📌 重点关注

  1. GoClick: Lightweight Element Grounding Model for Autonomous GUI Interaction | arXiv — 【重点关注】 Graphical User Interface (GUI) element grounding (precisely locating elements... 💡 轻量级GUI交互，为鸿蒙UI自动化提供新思路
  2. BARRED: Synthetic Training of Custom Policy Guardrails via Asymmetric Debate | arXiv — 【重点关注】 Deploying guardrails for custom policies remains challenging, as generic safe... 💡 辩论训练护栏，提升Agent安全性的关键方案
  3. Agentic Fusion of Large Atomic and Language Models to Accelerate Superconductors Discovery | arXiv — 【重点关注】 The discovery of novel materials is critical for global energy and quantum te... 💡 多模态Agent加速材料发现，AI科研范式创新

📋 其他值得关注

  1. GLM-5V-Turbo: Toward a Native Foundation Model for Multimodal Agents | arXiv — We present GLM-5V-Turbo, a step toward native foundation models for multimoda...
  2. Recursive Multi-Agent Systems | arXiv — Recursive or looped language models have recently emerged as a new scaling ax...
  3. AutoResearchBench: Benchmarking AI Agents on Complex Scientific Literature Discovery | arXiv — Autonomous scientific research is significantly advanced thanks to the develo...
  4. Large Language Models Explore by Latent Distilling | arXiv — Generating diverse responses is crucial for test-time scaling of large langua...
  5. DV-World: Benchmarking Data Visualization Agents in Real-World Scenarios | arXiv — Real-world data visualization (DV) requires native environmental grounding, c...
  6. ClawGym: A Scalable Framework for Building Effective Claw Agents | arXiv — Claw-style environments support multi-step workflows over local files, tools,...
  7. Programming with Data: Test-Driven Data Engineering for Self-Improving LLMs from Raw Corpora | arXiv — Reliably transferring specialized human knowledge from text into large langua...