AI论文速递 2026年05月27日（HuggingFace Daily Papers）

数据来源：https://huggingface.co/papers 采集时间：2026-05-27

📌 重点关注

1. Efficient Agentic Reinforcement Learning with On-Policy Intrinsic Knowledge Boundary Enhancement | arXiv — 【重点关注】 Agentic reinforcement learning (RL) has proven effective for training LLM-bas... 💡 移动端AI应用训练新范式
2. SEAL: Synergistic Co-Evolution of Agents and Learning Environments | arXiv — 【重点关注】 Large Language Model (LLM) agents are increasingly improved through interacti... 💡 Agent架构设计的重要参考
3. InstructSAM: Segment Any Instance with Any Instructions | arXiv — 【重点关注】 In this paper, we introduce InstructSAM, a unified and streamlined framework ... 💡 移动端图像分割优化方案

📋 其他值得关注

1. EvalVerse: Pipeline-Aware and Expert-Calibrated Benchmarking for Professional Cinematic Video Generation | arXiv — The rapid evolution of generative video foundation models has propelled the f...
2. Squeezing Capacity from Multimodal Large Language Models for Subject-driven Generation | arXiv — Subject-driven image generation aims to synthesize new images that preserve t...
3. Personalize-then-Store: Benchmarking and Learning Personalized Memory for Long-horizon Agents | arXiv — Existing large language model (LLM) based memory systems apply universal, sta...
4. Anticipate and Learn: Unleashing Idle-Time Compute in Proactive Agents | arXiv — While AI agents demonstrate remarkable capabilities in reasoning and tool use...
5. SimuWoB: Simulating Real-World Mobile Apps for Fast and Faithful GUI Agent Benchmarking | arXiv — Mobile GUI agents powered by large language models have progressed rapidly, c...
6. CUA-Gym: Scaling Verifiable Training Environments and Tasks for Computer-Use Agents | arXiv — Reinforcement learning with verifiable rewards (RLVR) has driven breakthrough...
7. Broadening Access to Transportation Safety Data with Generative AI: A Schema-Grounded Framework for Spatial Natural Language Queries | arXiv — Transportation safety analysis requires integrating crash records, roadway at...