{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:F5VCCCJFWVFTUSIPBBBLHLKRGS","short_pith_number":"pith:F5VCCCJF","schema_version":"1.0","canonical_sha256":"2f6a210925b54b3a490f0842b3ad513483f50d43556766044b7b726b5720f2f3","source":{"kind":"arxiv","id":"2602.08939","version":2},"attestation_state":"computed","paper":{"title":"CausalT5k: Diagnosing Refusal and Failure Modes in Trustworthy Causal Reasoning Across Causal Rungs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Andy Ouyang, Ankit Rai, Daphne Barretto, Edward Y. Chang, Gia Ancone, Longling Geng, Matthew John Hayes, Matthew Wolfman, Patrick Flanagan, Rachael Cooper, Sameer Vijay, Theodore Wu, Yuqiao Zeng","submitted_at":"2026-02-09T17:36:56Z","abstract_excerpt":"Large language models increasingly produce fluent causal explanations, yet they often fail in ways aggregate accuracy cannot diagnose: confusing association with intervention, abandoning correct judgments under pressure, over-refusing valid claims, or answering when evidence is underdetermined. We introduce CTK, a diagnostic benchmark of 5,147 cases and growing, across 10 domains and all three levels of Pearl's Ladder of Causation. Unlike benchmarks that only score correctness, CTK reveals why a model failed by annotating causal rung, trap type, pressure sensitivity, refusal quality, and Utili"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2602.08939","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-02-09T17:36:56Z","cross_cats_sorted":[],"title_canon_sha256":"e6a82b7a1ba4c1d2e413011b67ef3022d36b904d119a6f06235a4fc0b7b8d988","abstract_canon_sha256":"cea50e4118cd0dec5bf9d5166be712001adc7d62bf39dbee5613bb9a12287d98"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:09:54.748432Z","signature_b64":"vcmZDratEj+t4FlXneV5ArJ9IRltHs3RnZITP5nq5mq5CxTic5G/g8YqgmOVfwAUpiEF+kICGVTmyYhTD+QEAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2f6a210925b54b3a490f0842b3ad513483f50d43556766044b7b726b5720f2f3","last_reissued_at":"2026-06-19T16:09:54.748000Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:09:54.748000Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CausalT5k: Diagnosing Refusal and Failure Modes in Trustworthy Causal Reasoning Across Causal Rungs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Andy Ouyang, Ankit Rai, Daphne Barretto, Edward Y. Chang, Gia Ancone, Longling Geng, Matthew John Hayes, Matthew Wolfman, Patrick Flanagan, Rachael Cooper, Sameer Vijay, Theodore Wu, Yuqiao Zeng","submitted_at":"2026-02-09T17:36:56Z","abstract_excerpt":"Large language models increasingly produce fluent causal explanations, yet they often fail in ways aggregate accuracy cannot diagnose: confusing association with intervention, abandoning correct judgments under pressure, over-refusing valid claims, or answering when evidence is underdetermined. We introduce CTK, a diagnostic benchmark of 5,147 cases and growing, across 10 domains and all three levels of Pearl's Ladder of Causation. Unlike benchmarks that only score correctness, CTK reveals why a model failed by annotating causal rung, trap type, pressure sensitivity, refusal quality, and Utili"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.08939","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2602.08939/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2602.08939","created_at":"2026-06-19T16:09:54.748056+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.08939v2","created_at":"2026-06-19T16:09:54.748056+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.08939","created_at":"2026-06-19T16:09:54.748056+00:00"},{"alias_kind":"pith_short_12","alias_value":"F5VCCCJFWVFT","created_at":"2026-06-19T16:09:54.748056+00:00"},{"alias_kind":"pith_short_16","alias_value":"F5VCCCJFWVFTUSIP","created_at":"2026-06-19T16:09:54.748056+00:00"},{"alias_kind":"pith_short_8","alias_value":"F5VCCCJF","created_at":"2026-06-19T16:09:54.748056+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2605.30329","citing_title":"SoundnessBench: Can Your AI Scientist Really Tell Good Research Ideas from Bad Ones?","ref_index":3,"is_internal_anchor":true},{"citing_arxiv_id":"2605.09330","citing_title":"The Trap of Trajectory: Towards Understanding and Mitigating Spurious Correlations in Agentic Memory","ref_index":12,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/F5VCCCJFWVFTUSIPBBBLHLKRGS","json":"https://pith.science/pith/F5VCCCJFWVFTUSIPBBBLHLKRGS.json","graph_json":"https://pith.science/api/pith-number/F5VCCCJFWVFTUSIPBBBLHLKRGS/graph.json","events_json":"https://pith.science/api/pith-number/F5VCCCJFWVFTUSIPBBBLHLKRGS/events.json","paper":"https://pith.science/paper/F5VCCCJF"},"agent_actions":{"view_html":"https://pith.science/pith/F5VCCCJFWVFTUSIPBBBLHLKRGS","download_json":"https://pith.science/pith/F5VCCCJFWVFTUSIPBBBLHLKRGS.json","view_paper":"https://pith.science/paper/F5VCCCJF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.08939&json=true","fetch_graph":"https://pith.science/api/pith-number/F5VCCCJFWVFTUSIPBBBLHLKRGS/graph.json","fetch_events":"https://pith.science/api/pith-number/F5VCCCJFWVFTUSIPBBBLHLKRGS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/F5VCCCJFWVFTUSIPBBBLHLKRGS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/F5VCCCJFWVFTUSIPBBBLHLKRGS/action/storage_attestation","attest_author":"https://pith.science/pith/F5VCCCJFWVFTUSIPBBBLHLKRGS/action/author_attestation","sign_citation":"https://pith.science/pith/F5VCCCJFWVFTUSIPBBBLHLKRGS/action/citation_signature","submit_replication":"https://pith.science/pith/F5VCCCJFWVFTUSIPBBBLHLKRGS/action/replication_record"}},"created_at":"2026-06-19T16:09:54.748056+00:00","updated_at":"2026-06-19T16:09:54.748056+00:00"}