{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-08-21-003",
    "source_story_id": "tmp-story-scape-scenario-policy-evaluation",
    "edition_id": "mp-2026-08-21-morning-0043",
    "edition_url": "https://themachinepress.com/edition/2026-08-21",
    "position": 3,
    "story_type": "dispatch",
    "section": "robotics",
    "editorial_classification": "editorial",
    "headline": "The Average Hid Where the Robot Would Fail",
    "slug": "the-average-hid-where-the-robot-would-fail",
    "dek": "SCAPE predicted scenario-specific real-world policy performance from limited paired trials and large simulation runs.",
    "summary": "SCAPE predicted scenario-specific real-world policy performance from limited paired trials and large simulation runs.",
    "body_text": "SCAPE corrects simulation labels for sim-to-real bias, then uses conformal prediction to estimate real-world policy performance for particular deployment scenarios. In autonomous-driving and quadruped studies, the authors report lower scenario-level error than neural and aggregate baselines, narrower calibrated intervals and better out-of-distribution behavior. A physical Unitree Go2 test supports the evaluation method on one platform; it does not certify a policy for unrestricted deployment.",
    "why_it_matters": "SCAPE predicted scenario-specific real-world policy performance from limited paired trials and large simulation runs.",
    "limitations": [
      "A physical Unitree Go2 test supports the evaluation method on one platform; it does not certify a policy for unrestricted deployment."
    ],
    "importance": 9,
    "canonical_url": "https://themachinepress.com/story/mp-2026-08-21-003/the-average-hid-where-the-robot-would-fail",
    "json_url": "https://themachinepress.com/story/mp-2026-08-21-003.json",
    "first_published_at": "2026-08-21T09:00:00.000-04:00",
    "modified_at": "2026-08-21T09:00:00.000-04:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-08-21-003-001",
        "text": "SCAPE predicted scenario-specific real-world policy performance from limited paired trials and large simulation runs.",
        "source_ids": [
          "source-2026-08-21-003"
        ],
        "qualification": "A physical Unitree Go2 test supports the evaluation method on one platform; it does not certify a policy for unrestricted deployment."
      }
    ],
    "source_ids": [
      "source-2026-08-21-003"
    ],
    "tags": [
      "robot evaluation",
      "sim to real",
      "conformal prediction"
    ],
    "image_url": "https://themachinepress.com/issues/2026-08-21/scape-robot-file-image.webp",
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-08-21-003",
      "title": "arXiv preprint 2608.19425",
      "publisher": "arXiv",
      "url": "https://arxiv.org/abs/2608.19425",
      "canonical_url": "https://arxiv.org/abs/2608.19425",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-08-18T20:00:00.000-04:00",
      "accessed_at": "2026-08-21T08:22:06.000-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-08-21-003-001"
      ]
    }
  ],
  "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 Average Hid Where the Robot Would Fail",
    "publisher": "The Machine Press",
    "published_at": "2026-08-21T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-08-21-003/the-average-hid-where-the-robot-would-fail"
  }
}