{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:ATKZT7D2XSNXZIWRFYLIZA6W5D","short_pith_number":"pith:ATKZT7D2","schema_version":"1.0","canonical_sha256":"04d599fc7abc9b7ca2d12e168c83d6e8ebb30ff525549ecb5fbc62c32f373fdd","source":{"kind":"arxiv","id":"2410.07432","version":2},"attestation_state":"computed","paper":{"title":"Can Transformers Reason Logically? A Study in SAT Solving","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LO"],"primary_cat":"cs.LG","authors_text":"Chris Esposo, Jacob Abernethy, Leyan Pan, Vijay Ganesh, Wenke Lee","submitted_at":"2024-10-09T21:01:52Z","abstract_excerpt":"We formally study the logical reasoning capabilities of decoder-only Transformers in the context of the boolean satisfiability (SAT) problem. First, we prove by construction that decoder-only Transformers can decide 3-SAT, in a non-uniform model of computation, using backtracking and deduction via Chain-of-Thought (CoT). %We prove its correctness by showing trace equivalence to the well-known DPLL SAT-solving algorithm. Second, we implement our construction as a PyTorch model with a tool (PARAT) that we designed to empirically demonstrate its correctness and investigate its properties. Third, "},"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":"2410.07432","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-10-09T21:01:52Z","cross_cats_sorted":["cs.AI","cs.LO"],"title_canon_sha256":"1da22318f295f8046614a18f2c672578f49790ceb855d196b42bccf7868a32e4","abstract_canon_sha256":"36ba8dd34c5f7b1f941965bae2eb7337073472af08448f862784892d2965bc96"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:11:25.589880Z","signature_b64":"JkO97ymdNgRqJjw5pOFc6gEBay/RkeyOBUudFLrUISPsSIoaNjeKFdiBZEBiicV0dR7ZkQ1pE866Ljl4bGM7AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"04d599fc7abc9b7ca2d12e168c83d6e8ebb30ff525549ecb5fbc62c32f373fdd","last_reissued_at":"2026-07-05T10:11:25.589361Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:11:25.589361Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Can Transformers Reason Logically? A Study in SAT Solving","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LO"],"primary_cat":"cs.LG","authors_text":"Chris Esposo, Jacob Abernethy, Leyan Pan, Vijay Ganesh, Wenke Lee","submitted_at":"2024-10-09T21:01:52Z","abstract_excerpt":"We formally study the logical reasoning capabilities of decoder-only Transformers in the context of the boolean satisfiability (SAT) problem. First, we prove by construction that decoder-only Transformers can decide 3-SAT, in a non-uniform model of computation, using backtracking and deduction via Chain-of-Thought (CoT). %We prove its correctness by showing trace equivalence to the well-known DPLL SAT-solving algorithm. Second, we implement our construction as a PyTorch model with a tool (PARAT) that we designed to empirically demonstrate its correctness and investigate its properties. Third, "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.07432","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/2410.07432/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":"2410.07432","created_at":"2026-07-05T10:11:25.589426+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.07432v2","created_at":"2026-07-05T10:11:25.589426+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.07432","created_at":"2026-07-05T10:11:25.589426+00:00"},{"alias_kind":"pith_short_12","alias_value":"ATKZT7D2XSNX","created_at":"2026-07-05T10:11:25.589426+00:00"},{"alias_kind":"pith_short_16","alias_value":"ATKZT7D2XSNXZIWR","created_at":"2026-07-05T10:11:25.589426+00:00"},{"alias_kind":"pith_short_8","alias_value":"ATKZT7D2","created_at":"2026-07-05T10:11:25.589426+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.16632","citing_title":"Learning How to Cube","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2511.02521","citing_title":"Large Lemma Miners: Can LLMs do Induction Proofs for Hardware?","ref_index":12,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ATKZT7D2XSNXZIWRFYLIZA6W5D","json":"https://pith.science/pith/ATKZT7D2XSNXZIWRFYLIZA6W5D.json","graph_json":"https://pith.science/api/pith-number/ATKZT7D2XSNXZIWRFYLIZA6W5D/graph.json","events_json":"https://pith.science/api/pith-number/ATKZT7D2XSNXZIWRFYLIZA6W5D/events.json","paper":"https://pith.science/paper/ATKZT7D2"},"agent_actions":{"view_html":"https://pith.science/pith/ATKZT7D2XSNXZIWRFYLIZA6W5D","download_json":"https://pith.science/pith/ATKZT7D2XSNXZIWRFYLIZA6W5D.json","view_paper":"https://pith.science/paper/ATKZT7D2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.07432&json=true","fetch_graph":"https://pith.science/api/pith-number/ATKZT7D2XSNXZIWRFYLIZA6W5D/graph.json","fetch_events":"https://pith.science/api/pith-number/ATKZT7D2XSNXZIWRFYLIZA6W5D/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ATKZT7D2XSNXZIWRFYLIZA6W5D/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ATKZT7D2XSNXZIWRFYLIZA6W5D/action/storage_attestation","attest_author":"https://pith.science/pith/ATKZT7D2XSNXZIWRFYLIZA6W5D/action/author_attestation","sign_citation":"https://pith.science/pith/ATKZT7D2XSNXZIWRFYLIZA6W5D/action/citation_signature","submit_replication":"https://pith.science/pith/ATKZT7D2XSNXZIWRFYLIZA6W5D/action/replication_record"}},"created_at":"2026-07-05T10:11:25.589426+00:00","updated_at":"2026-07-05T10:11:25.589426+00:00"}