{
  "id": 12318960,
  "title": "Recursive Video In-Context Learning for Agentic Robot",
  "url": "https://urgent.news/2026/10/05/recursive-video-in-context-learning-for-agentic-robot",
  "topic": "ai",
  "section": "AI",
  "published": "2026-10-05T17:59:11.000Z",
  "source": {
    "name": "arXiv cs.AI",
    "slug": "arxiv-cs-ai",
    "url": "https://arxiv.org/abs/2610.06843v1"
  },
  "original_language": "en",
  "account": null,
  "summary": "LLM agents that orchestrate frozen vision-language-action (VLA) policies improve across episodes through text memory, which records what the agent did but not how the task is done. A demonstration video shows it, but fits poorly into an agent's context. The full video slows every turn, fixed keyframes lose the contact detail that decides whether a grasp holds, and what the agent needs shifts from…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}