{
  "id": 1479,
  "url": "https://arxiv.org/abs/2606.05395v1",
  "title": "VASO: Formally Verifiable Self-Evolving Skills for Physical AI Agents",
  "summary": "Reusable robot skills are becoming the basic units through which embodied agents turn open-ended instructions into long-horizon physical behavior. We argue that, while foundation models have collapsed the cost of creating these skills, the cost of trusting them has not. Existing skill-evolution loops refine skills through execution feedback, unit tests, environment reward, or LLM self-critique, but these signals provide only trace-level evidence: they show that a skill worked on sampled executio",
  "authors": "Yunhao Yang, Neel P. Bhatt, Kevin Wang, Samuel Tetteh, Zhangyang Wang, Ufuk Topcu",
  "category": "research",
  "topics": "agents-autonomy,environment",
  "orgs": null,
  "regions": null,
  "published_at": "2026-06-03T20:02:35.000Z",
  "fetched_at": "2026-07-14T14:15:17.104Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/1479",
  "original_url": "https://arxiv.org/abs/2606.05395v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}