Daily AI Papers Briefing (2026-06-01)

ずんだもんのHugging Faceニュース · May 31, 2026 · 3m

【本日の論文】

1. AgentDoG 1.5: A Lightweight and Scalable Alignment Framework for AI Agent Safety and Security

https://huggingface.co/papers/2605.29801

2. Qwen-VLA: Unifying Vision-Language-Action Modeling across Tasks, Environments, and Robot Embodiments

https://huggingface.co/papers/2605.30280

3. OmniRetrieval: Unified Retrieval across Heterogeneous Knowledge Sources

https://huggingface.co/papers/2605.29250

4. CollectionLoRA: Collecting 50 Effects in 1 LoRA via Multi-Teacher On-Policy Distillation

https://huggingface.co/papers/2605.25378

5. minWM: A Full-Stack Open-Source Framework for Real-Time…