{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:NADE2IOJTMO5EKISCMAMYCSUVG","short_pith_number":"pith:NADE2IOJ","schema_version":"1.0","canonical_sha256":"68064d21c99b1dd229121300cc0a54a9b76633df1e00d198a5f089d2c536cf39","source":{"kind":"arxiv","id":"2507.23488","version":1},"attestation_state":"computed","paper":{"title":"Causal Reasoning in Pieces: Modular In-Context Learning for Causal Discovery","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Kacper Kadziolka, Saber Salehkaleybar","submitted_at":"2025-07-31T12:10:27Z","abstract_excerpt":"Causal inference remains a fundamental challenge for large language models. Recent advances in internal reasoning with large language models have sparked interest in whether state-of-the-art reasoning models can robustly perform causal discovery-a task where conventional models often suffer from severe overfitting and near-random performance under data perturbations. We study causal discovery on the Corr2Cause benchmark using the emergent OpenAI's o-series and DeepSeek-R model families and find that these reasoning-first architectures achieve significantly greater native gains than prior appro"},"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":"2507.23488","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-07-31T12:10:27Z","cross_cats_sorted":[],"title_canon_sha256":"ded71e280d6d2a9560cfff31c9104c81569fe2df775d8ef791a74bd01c372f00","abstract_canon_sha256":"cd75df3296e036533dc30d3305db52975547f41f8e9bef782f49850e327aef5a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:46:25.231679Z","signature_b64":"Dl7hpFPgBRs0h2ifvQiip9AN0XYMppD/v064iQpG7NxmmiR4FRgI4KNPu+LntpQhjvfyrOkFcMDmSquXXJdKAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"68064d21c99b1dd229121300cc0a54a9b76633df1e00d198a5f089d2c536cf39","last_reissued_at":"2026-07-05T11:46:25.231166Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:46:25.231166Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Causal Reasoning in Pieces: Modular In-Context Learning for Causal Discovery","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Kacper Kadziolka, Saber Salehkaleybar","submitted_at":"2025-07-31T12:10:27Z","abstract_excerpt":"Causal inference remains a fundamental challenge for large language models. Recent advances in internal reasoning with large language models have sparked interest in whether state-of-the-art reasoning models can robustly perform causal discovery-a task where conventional models often suffer from severe overfitting and near-random performance under data perturbations. We study causal discovery on the Corr2Cause benchmark using the emergent OpenAI's o-series and DeepSeek-R model families and find that these reasoning-first architectures achieve significantly greater native gains than prior appro"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.23488","kind":"arxiv","version":1},"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/2507.23488/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":"2507.23488","created_at":"2026-07-05T11:46:25.231227+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.23488v1","created_at":"2026-07-05T11:46:25.231227+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.23488","created_at":"2026-07-05T11:46:25.231227+00:00"},{"alias_kind":"pith_short_12","alias_value":"NADE2IOJTMO5","created_at":"2026-07-05T11:46:25.231227+00:00"},{"alias_kind":"pith_short_16","alias_value":"NADE2IOJTMO5EKIS","created_at":"2026-07-05T11:46:25.231227+00:00"},{"alias_kind":"pith_short_8","alias_value":"NADE2IOJ","created_at":"2026-07-05T11:46:25.231227+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.27567","citing_title":"Why LLMs Fail at Causal Discovery and How Interventional Agents Escape","ref_index":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NADE2IOJTMO5EKISCMAMYCSUVG","json":"https://pith.science/pith/NADE2IOJTMO5EKISCMAMYCSUVG.json","graph_json":"https://pith.science/api/pith-number/NADE2IOJTMO5EKISCMAMYCSUVG/graph.json","events_json":"https://pith.science/api/pith-number/NADE2IOJTMO5EKISCMAMYCSUVG/events.json","paper":"https://pith.science/paper/NADE2IOJ"},"agent_actions":{"view_html":"https://pith.science/pith/NADE2IOJTMO5EKISCMAMYCSUVG","download_json":"https://pith.science/pith/NADE2IOJTMO5EKISCMAMYCSUVG.json","view_paper":"https://pith.science/paper/NADE2IOJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.23488&json=true","fetch_graph":"https://pith.science/api/pith-number/NADE2IOJTMO5EKISCMAMYCSUVG/graph.json","fetch_events":"https://pith.science/api/pith-number/NADE2IOJTMO5EKISCMAMYCSUVG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NADE2IOJTMO5EKISCMAMYCSUVG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NADE2IOJTMO5EKISCMAMYCSUVG/action/storage_attestation","attest_author":"https://pith.science/pith/NADE2IOJTMO5EKISCMAMYCSUVG/action/author_attestation","sign_citation":"https://pith.science/pith/NADE2IOJTMO5EKISCMAMYCSUVG/action/citation_signature","submit_replication":"https://pith.science/pith/NADE2IOJTMO5EKISCMAMYCSUVG/action/replication_record"}},"created_at":"2026-07-05T11:46:25.231227+00:00","updated_at":"2026-07-05T11:46:25.231227+00:00"}