{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-08-19-026",
    "source_story_id": "tmp-story-probe-manipulation-grounded-vqa",
    "edition_id": "mp-2026-08-19-morning-0041",
    "edition_url": "https://themachinepress.com/edition/2026-08-19",
    "position": 15,
    "story_type": "dispatch",
    "section": "robotics",
    "editorial_classification": "editorial",
    "headline": "The Robot Moved the Clutter Before It Answered",
    "slug": "the-robot-moved-the-clutter-before-it-answered",
    "dek": "PROBE turns visual question answering into an active task where a robot must reveal occluded objects before deciding.",
    "summary": "PROBE turns visual question answering into an active task where a robot must reveal occluded objects before deciding.",
    "body_text": "PROBE-Bench contains 150 simulated tabletop tasks across six question types, with picking and pushing tools that change the scene between observations. Agentic tool use outperformed perception-only baselines by 8.0 percent on average, and distilled open-weight agents gained another 11.5 percent over their off-the-shelf agent baselines. The authors also report real-tabletop transfer, but the benchmark remains a bounded step toward robots that can answer questions about hidden household state.",
    "why_it_matters": "PROBE turns visual question answering into an active task where a robot must reveal occluded objects before deciding.",
    "limitations": [
      "Agentic tool use outperformed perception-only baselines by 8.0 percent on average, and distilled open-weight agents gained another 11.5 percent over their off-the-shelf agent baselines."
    ],
    "importance": 9,
    "canonical_url": "https://themachinepress.com/story/mp-2026-08-19-026/the-robot-moved-the-clutter-before-it-answered",
    "json_url": "https://themachinepress.com/story/mp-2026-08-19-026.json",
    "first_published_at": "2026-08-19T09:00:00.000-04:00",
    "modified_at": "2026-08-19T09:00:00.000-04:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-08-19-026-001",
        "text": "PROBE turns visual question answering into an active task where a robot must reveal occluded objects before deciding.",
        "source_ids": [
          "source-2026-08-19-015"
        ],
        "qualification": "Agentic tool use outperformed perception-only baselines by 8.0 percent on average, and distilled open-weight agents gained another 11.5 percent over their off-the-shelf agent baselines."
      }
    ],
    "source_ids": [
      "source-2026-08-19-015"
    ],
    "tags": [
      "visual question answering",
      "robot manipulation",
      "occlusion",
      "VLM agents"
    ],
    "image_url": null,
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-08-19-015",
      "title": "arXiv preprint 2608.17129",
      "publisher": "arXiv",
      "url": "https://arxiv.org/abs/2608.17129",
      "canonical_url": "https://arxiv.org/abs/2608.17129",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-08-16T20:00:00.000-04:00",
      "accessed_at": "2026-08-19T08:22:00.000-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-08-19-026-001"
      ]
    }
  ],
  "corrections": [],
  "publisher": {
    "name": "The Machine Press",
    "url": "https://themachinepress.com",
    "description": "A daily newspaper for the age of artificial intelligence."
  },
  "cite_this_report": {
    "title": "The Robot Moved the Clutter Before It Answered",
    "publisher": "The Machine Press",
    "published_at": "2026-08-19T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-08-19-026/the-robot-moved-the-clutter-before-it-answered"
  }
}