{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-10-16-016",
    "source_story_id": "tmp-brief-mindeval2-mental-health-support",
    "edition_id": "mp-2026-10-16-morning-0098",
    "edition_url": "https://themachinepress.com/edition/2026-10-16",
    "position": 18,
    "story_type": "ticker",
    "section": "benchmarks-evals",
    "editorial_classification": "editorial",
    "headline": "Mental-Health Evaluation Finds Guidance Calibration Still Fails",
    "slug": "mental-health-evaluation-finds-guidance-calibration-still-fails",
    "dek": "MindEval2 combines psychologist-authored criteria with harder synthetic patients and longitudinal memory.",
    "summary": "MindEval2 combines psychologist-authored criteria with harder synthetic patients and longitudinal memory.",
    "body_text": "Testing 18 proprietary and open-weight models, the authors found that surface quality and boundary behavior improved while guidance calibration remained the dominant source of the quality gap. The framework is an evaluation tool, not clinical validation or a substitute for professional care.",
    "why_it_matters": "MindEval2 combines psychologist-authored criteria with harder synthetic patients and longitudinal memory.",
    "limitations": [
      "The framework is an evaluation tool, not clinical validation or a substitute for professional care."
    ],
    "importance": 7,
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-16-016/mental-health-evaluation-finds-guidance-calibration-still-fails",
    "json_url": "https://themachinepress.com/story/mp-2026-10-16-016.json",
    "first_published_at": "2026-10-16T09:00:00.000-04:00",
    "modified_at": "2026-10-16T09:00:00.000-04:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-10-16-016-001",
        "text": "MindEval2 combines psychologist-authored criteria with harder synthetic patients and longitudinal memory.",
        "source_ids": [
          "source-2026-10-16-018"
        ],
        "qualification": "The framework is an evaluation tool, not clinical validation or a substitute for professional care."
      }
    ],
    "source_ids": [
      "source-2026-10-16-018"
    ],
    "tags": [
      "mental health",
      "evaluation",
      "language models"
    ],
    "image_url": null,
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-10-16-018",
      "title": "arXiv preprint 2610.22314, original v1",
      "publisher": "arXiv / submitting research authors",
      "url": "https://arxiv.org/abs/2610.22314v1",
      "canonical_url": "https://arxiv.org/abs/2610.22314v1",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-10-15T13:51:01.000-04:00",
      "accessed_at": "2026-10-16T08:26:36.400-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-10-16-016-001"
      ]
    }
  ],
  "corrections": [],
  "publisher": {
    "name": "The Machine Press",
    "url": "https://themachinepress.com",
    "description": "A daily newspaper for the age of artificial intelligence."
  },
  "cite_this_report": {
    "title": "Mental-Health Evaluation Finds Guidance Calibration Still Fails",
    "publisher": "The Machine Press",
    "published_at": "2026-10-16T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-16-016/mental-health-evaluation-finds-guidance-calibration-still-fails"
  }
}