{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-11-05-009",
    "source_story_id": "tmp-story-generalist-robot-rl",
    "edition_id": "mp-2026-11-05-morning-0118",
    "edition_url": "https://themachinepress.com/edition/2026-11-05",
    "position": 9,
    "story_type": "dispatch",
    "section": "robotics",
    "editorial_classification": "editorial",
    "headline": "A Robot Policy Improves Beyond the Tasks That Trained It",
    "slug": "a-robot-policy-improves-beyond-the-tasks-that-trained-it",
    "dek": "A structured study finds online reinforcement learning can raise performance across many deployment tasks and sometimes transfer gains to held-out tasks.",
    "summary": "A structured study finds online reinforcement learning can raise performance across many deployment tasks and sometimes transfer gains to held-out tasks.",
    "body_text": "The authors vary value estimation, policy extraction, resource allocation and the training-task distribution around a generalist robot foundation policy. Their results show simultaneous improvement across diverse tasks, with gains on tasks excluded from online training under the right recipe. Value-based reinforcement learning and a sufficiently broad training distribution are central to that transfer. The work provides evidence for multi-task online improvement in the evaluated system; it does not show open-ended self-improvement or guaranteed transfer to arbitrary robots.",
    "why_it_matters": "A structured study finds online reinforcement learning can raise performance across many deployment tasks and sometimes transfer gains to held-out tasks.",
    "limitations": [
      "The work provides evidence for multi-task online improvement in the evaluated system; it does not show open-ended self-improvement or guaranteed transfer to arbitrary robots."
    ],
    "importance": 9,
    "canonical_url": "https://themachinepress.com/story/mp-2026-11-05-009/a-robot-policy-improves-beyond-the-tasks-that-trained-it",
    "json_url": "https://themachinepress.com/story/mp-2026-11-05-009.json",
    "first_published_at": "2026-11-05T09:00:00.000-05:00",
    "modified_at": "2026-11-05T09:00:00.000-05:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-11-05-009-001",
        "text": "A structured study finds online reinforcement learning can raise performance across many deployment tasks and sometimes transfer gains to held-out tasks.",
        "source_ids": [
          "source-2026-11-05-009"
        ],
        "qualification": "The work provides evidence for multi-task online improvement in the evaluated system; it does not show open-ended self-improvement or guaranteed transfer to arbitrary robots."
      }
    ],
    "source_ids": [
      "source-2026-11-05-009"
    ],
    "tags": [
      "robot foundation models",
      "reinforcement learning",
      "online adaptation"
    ],
    "image_url": null,
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-11-05-009",
      "title": "arXiv preprint 2611.06905, original v1",
      "publisher": "arXiv",
      "url": "https://arxiv.org/abs/2611.06905v1",
      "canonical_url": "https://arxiv.org/abs/2611.06905v1",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-11-04T13:52:28.000-05:00",
      "accessed_at": "2026-11-05T08:29:24.413-05:00",
      "supports_claim_ids": [
        "claim-mp-2026-11-05-009-001"
      ]
    }
  ],
  "corrections": [],
  "publisher": {
    "name": "The Machine Press",
    "url": "https://themachinepress.com",
    "description": "A daily newspaper for the age of artificial intelligence."
  },
  "cite_this_report": {
    "title": "A Robot Policy Improves Beyond the Tasks That Trained It",
    "publisher": "The Machine Press",
    "published_at": "2026-11-05T09:00:00.000-05:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-11-05-009/a-robot-policy-improves-beyond-the-tasks-that-trained-it"
  }
}