Step 2: tool results

get_claim

{
  "include": [
    "arguments",
    "subclaims",
    "related"
  ],
  "claim_id": "846a4d41-3752-45ed-9210-b4177d75b5e9"
}

{"claim":{"id":"846a4d41-3752-45ed-9210-b4177d75b5e9","kind":"claim","canonical_form":"In API-only model diffing, an improved LLM baseline performs comparably to SAE-based methods and surfaces more abstract behavioral differences.","claim_type":"empirical_derived","state":"active","decomposition_status":"complete","importance":0.5,"contestation":0.45,"domains":[],"tags":[{"slug":"mechanistic-interpretability","name":"Mechanistic Interpretability"},{"slug":"large-language-models","name":"Large language models"},{"slug":"sparse-autoencoders","name":"Sparse Autoencoders"}],"steward_state":"running","stewarded_at":"2026-11-14T20:43:11.264Z","canonical_direction_note":"The affirmative form states the source's comparative finding about an improved API-only LLM baseline and SAE-based model-diffing methods.","created_at":"2026-11-13T22:50:23.580Z","updated_at":"2026-11-14T20:39:57.467Z","page_url":"https://minerval.ai/claims/846a4d41-3752-45ed-9210-b4177d75b5e9"},"assessment":null,"coverage":{"examined":false,"on":[],"note":"No judgment of this node is in force. That is not a finding either way."},"quantity":null,"threshold_of":null,"question":null,"formalization":null,"formalization_pending":null,"formalization_history":[],"verification":null,"lean_checks":[],"lean_checks_total":0,"waiting_messages":[{"trigger":"structure_and_assess","from":"the system","sender_node_id":null,"sent_at":"2026-11-13T22:50:25.216Z","paid":false,"summary":"New claim onboarded. Give it the structure and assessment its importance warrants; a minor or settled claim may need no decomposition."},{"trigger":"mandate_plan","from":"a mandate's plan","sender_node_id":null,"sent_at":"2026-11-14T20:39:57.468Z","paid":true,"summary":"First assessment. Read the model-diffing comparison behind this claim and the SAE-based method it is compared with. Establish: which model pairs, how differences were scored and by whom, what the impr…"}],"requests":[],"requests_total":0,"subclaim_count":0,"instance_count":1,"source_map":null,"subclaims":[],"subclaims_assessed":0,"arguments":[],"related":[],"unplaced_argument_passages":[{"instance_id":"0bffbda7-9cdc-424b-ae78-a76d5a0296e2","source":{"id":"a5178d56-97c3-499f-9ce3-1d7fe1dad10e","url":"https://arxiv.org/pdf/2602.10371","title":"Simple LLM Baselines are Competitive for Model Diffing"},"verbatim_text":"Across all three experiments, accuracy and frequency are comparable between the diffing meth-\nods (Figure 2), but the LLM-based approach consistently produces more abstract hypotheses.","context":"Results summary across the three experimental setups","stance":"advances","direction":"for"},{"instance_id":"c0f0f8e8-0dc5-48fa-90fa-399cd2787b7a","source":{"id":"a5178d56-97c3-499f-9ce3-1d7fe1dad10e","url":"https://arxiv.org/pdf/2602.10371","title":"Simple LLM Baselines are Competitive for Model Diffing"},"verbatim_text":"LLM distributions are concentrated at high abstraction levels\n(means between 3.8 and 4.3) with low variance, while SAE distributions are lower (means between\n2.4 and 3.5) with higher variance.","context":"Abstraction-level distributions across Qwen, Gemma, and Gemini experiments","stance":"advances","direction":"for"}],"unplaced_argument_passages_total":2}

get_claim_instances

{
  "claim_id": "846a4d41-3752-45ed-9210-b4177d75b5e9"
}

{
  "claim_id": "846a4d41-3752-45ed-9210-b4177d75b5e9",
  "total": 1,
  "limit": 20,
  "offset": 0,
  "has_more": false,
  "instances": [
    {
      "id": "e6c04843-55d5-4fdd-8b7c-5fb601be2508",
      "source_id": "a5178d56-97c3-499f-9ce3-1d7fe1dad10e",
      "verbatim_text": "Our results show that an improved\nLLM-based baseline performs comparably to the SAE-based method while typi-\ncally surfacing more abstract behavioral differences.",
      "context": "Abstract and intro, summarising a comparison of API-only model diffing methods across three setups (Qwen emergent misalignment, Gemma hidden user-gender assumption, Gemini 2.5 Flash Lite revisions).",
      "stance": "affirms",
      "confidence": 0.85,
      "speaker": null,
      "publication": null,
      "source_date": null,
      "link": null,
      "source_title": "Simple LLM Baselines are Competitive for Model Diffing",
      "source_type": "preprint",
      "source_url": "https://arxiv.org/pdf/2602.10371",
      "source_published_date": "2026-02-10",
      "source_authors": [
        "Kempf, Elias",
        "Schrodi, Simon",
        "Cywiński, Bartosz",
        "Brox, Thomas",
        "Nanda, Neel",
        "Conmy, Arthur"
      ],
      "source_publisher": "arXiv.org",
      "source_doi": null,
      "source_interest": null,
      "recorded_by": "extractor",
      "display_rank": null,
      "reading": null,
      "draws_on": [],
      "story": {
        "standing": "untraced",
        "basis": "none",
        "depth": 0,
        "diverges": false,
        "downstream_total": 0
      }
    }
  ]
}