{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:VTYMM4MIPE7Q3QTSNTAG64SVVG","short_pith_number":"pith:VTYMM4MI","schema_version":"1.0","canonical_sha256":"acf0c67188793f0dc2726cc06f7255a99bd3d4b209f17fd1ab1643ef09425c22","source":{"kind":"arxiv","id":"2410.03767","version":2},"attestation_state":"computed","paper":{"title":"Reasoning Elicitation in Language Models via Counterfactual Feedback","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Aditya V. Nori, Alihan H\\\"uy\\\"uk, Jacqueline Maasch, Javier Gonz\\'alez, Xinnuo Xu","submitted_at":"2024-10-02T15:33:30Z","abstract_excerpt":"Despite the increasing effectiveness of language models, their reasoning capabilities remain underdeveloped. In particular, causal reasoning through counterfactual question answering is lacking. This work aims to bridge this gap. We first derive novel metrics that balance accuracy in factual and counterfactual questions, capturing a more complete view of the reasoning abilities of language models than traditional factual-only based metrics. Second, we propose several fine-tuning approaches that aim to elicit better reasoning mechanisms, in the sense of the proposed metrics. Finally, we evaluat"},"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":"2410.03767","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-02T15:33:30Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"eebaeb2a3e475dc352715434ccf1eceb508ca548ffc72603c41b96d7166ea9f9","abstract_canon_sha256":"cc009e85fe56f8b9cdd875d13f1eb35586c425d49907e9a08d64bf41ab3268b5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:31:47.690563Z","signature_b64":"ejcjt/VaoGOrr9YsI7e6MUy7MVahWlzo9f3V7SSR3tkSWFblqK9lbP8Uj2JxOjd9ayaMxXxEZEa6jutM0WIVAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"acf0c67188793f0dc2726cc06f7255a99bd3d4b209f17fd1ab1643ef09425c22","last_reissued_at":"2026-07-05T10:31:47.690017Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:31:47.690017Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reasoning Elicitation in Language Models via Counterfactual Feedback","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Aditya V. Nori, Alihan H\\\"uy\\\"uk, Jacqueline Maasch, Javier Gonz\\'alez, Xinnuo Xu","submitted_at":"2024-10-02T15:33:30Z","abstract_excerpt":"Despite the increasing effectiveness of language models, their reasoning capabilities remain underdeveloped. In particular, causal reasoning through counterfactual question answering is lacking. This work aims to bridge this gap. We first derive novel metrics that balance accuracy in factual and counterfactual questions, capturing a more complete view of the reasoning abilities of language models than traditional factual-only based metrics. Second, we propose several fine-tuning approaches that aim to elicit better reasoning mechanisms, in the sense of the proposed metrics. Finally, we evaluat"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.03767","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/2410.03767/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":"2410.03767","created_at":"2026-07-05T10:31:47.690079+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.03767v2","created_at":"2026-07-05T10:31:47.690079+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.03767","created_at":"2026-07-05T10:31:47.690079+00:00"},{"alias_kind":"pith_short_12","alias_value":"VTYMM4MIPE7Q","created_at":"2026-07-05T10:31:47.690079+00:00"},{"alias_kind":"pith_short_16","alias_value":"VTYMM4MIPE7Q3QTS","created_at":"2026-07-05T10:31:47.690079+00:00"},{"alias_kind":"pith_short_8","alias_value":"VTYMM4MI","created_at":"2026-07-05T10:31:47.690079+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.09079","citing_title":"CauSim: Scaling Causal Reasoning with Increasingly Complex Causal Simulators","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2604.10693","citing_title":"FACT-E: Causality-Inspired Evaluation for Trustworthy Chain-of-Thought Reasoning","ref_index":12,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VTYMM4MIPE7Q3QTSNTAG64SVVG","json":"https://pith.science/pith/VTYMM4MIPE7Q3QTSNTAG64SVVG.json","graph_json":"https://pith.science/api/pith-number/VTYMM4MIPE7Q3QTSNTAG64SVVG/graph.json","events_json":"https://pith.science/api/pith-number/VTYMM4MIPE7Q3QTSNTAG64SVVG/events.json","paper":"https://pith.science/paper/VTYMM4MI"},"agent_actions":{"view_html":"https://pith.science/pith/VTYMM4MIPE7Q3QTSNTAG64SVVG","download_json":"https://pith.science/pith/VTYMM4MIPE7Q3QTSNTAG64SVVG.json","view_paper":"https://pith.science/paper/VTYMM4MI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.03767&json=true","fetch_graph":"https://pith.science/api/pith-number/VTYMM4MIPE7Q3QTSNTAG64SVVG/graph.json","fetch_events":"https://pith.science/api/pith-number/VTYMM4MIPE7Q3QTSNTAG64SVVG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VTYMM4MIPE7Q3QTSNTAG64SVVG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VTYMM4MIPE7Q3QTSNTAG64SVVG/action/storage_attestation","attest_author":"https://pith.science/pith/VTYMM4MIPE7Q3QTSNTAG64SVVG/action/author_attestation","sign_citation":"https://pith.science/pith/VTYMM4MIPE7Q3QTSNTAG64SVVG/action/citation_signature","submit_replication":"https://pith.science/pith/VTYMM4MIPE7Q3QTSNTAG64SVVG/action/replication_record"}},"created_at":"2026-07-05T10:31:47.690079+00:00","updated_at":"2026-07-05T10:31:47.690079+00:00"}