{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:DTINKQS5IM2L4MFEVRCWMM2Z3H","short_pith_number":"pith:DTINKQS5","canonical_record":{"source":{"id":"2605.28602","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-27T15:18:45Z","cross_cats_sorted":["cs.CL","cs.LO"],"title_canon_sha256":"635831c300b28016780559cfd917ac4e49ce998627aad05ae67a6ac4e7860867","abstract_canon_sha256":"acf18ca6d3c4aefe1e56451907ee9e9adf9aa89b803975eec9e55ab3d656d5a8"},"schema_version":"1.0"},"canonical_sha256":"1cd0d5425d4334be30a4ac45663359d9de6a574401d032c64049dda8c3bcaec2","source":{"kind":"arxiv","id":"2605.28602","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.28602","created_at":"2026-05-28T02:04:57Z"},{"alias_kind":"arxiv_version","alias_value":"2605.28602v1","created_at":"2026-05-28T02:04:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.28602","created_at":"2026-05-28T02:04:57Z"},{"alias_kind":"pith_short_12","alias_value":"DTINKQS5IM2L","created_at":"2026-05-28T02:04:57Z"},{"alias_kind":"pith_short_16","alias_value":"DTINKQS5IM2L4MFE","created_at":"2026-05-28T02:04:57Z"},{"alias_kind":"pith_short_8","alias_value":"DTINKQS5","created_at":"2026-05-28T02:04:57Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:DTINKQS5IM2L4MFEVRCWMM2Z3H","target":"record","payload":{"canonical_record":{"source":{"id":"2605.28602","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-27T15:18:45Z","cross_cats_sorted":["cs.CL","cs.LO"],"title_canon_sha256":"635831c300b28016780559cfd917ac4e49ce998627aad05ae67a6ac4e7860867","abstract_canon_sha256":"acf18ca6d3c4aefe1e56451907ee9e9adf9aa89b803975eec9e55ab3d656d5a8"},"schema_version":"1.0"},"canonical_sha256":"1cd0d5425d4334be30a4ac45663359d9de6a574401d032c64049dda8c3bcaec2","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-28T02:04:57.511732Z","signature_b64":"zEyLMOSpAEArZOs3zfLnU/B1w4KlDXjt+Bh86sWPclHuQSDj1INheb5XLHjfbbAHcsWah6U0TN6C2X0GMF5KCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1cd0d5425d4334be30a4ac45663359d9de6a574401d032c64049dda8c3bcaec2","last_reissued_at":"2026-05-28T02:04:57.511289Z","signature_status":"signed_v1","first_computed_at":"2026-05-28T02:04:57.511289Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.28602","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-28T02:04:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"oNu34F+ggprjuI50WLKDPLVhsGDu6BVAPSzYDDgnErxZZn58bdnWlWXlxP/BuUTI9BJkEz+U36fntNiaTVkXDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-30T17:27:09.431892Z"},"content_sha256":"83c7f58bfc68ac559be6ba8a7d156390542af09d9b98684d230fd2c8e106244e","schema_version":"1.0","event_id":"sha256:83c7f58bfc68ac559be6ba8a7d156390542af09d9b98684d230fd2c8e106244e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:DTINKQS5IM2L4MFEVRCWMM2Z3H","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Satisfiability Solving with LLMs: A Matched-Pair Evaluation of Reasoning Capability","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LO"],"primary_cat":"cs.AI","authors_text":"Leizhen Zhang, Sheng Chen, Shuhan Chen","submitted_at":"2026-05-27T15:18:45Z","abstract_excerpt":"Large language models (LLMs) are increasingly used for tasks that implicitly reduce to Boolean satisfiability (SAT), yet their reasoning ability on SAT remains unclear. We present a systematic study of LLMs on 2-SAT and 3-SAT, together with two canonical reductions, Vertex Cover and discrete 3D packing, to probe representation-invariant reasoning. We first evaluate models using conventional metrics, including accuracy, precision, recall, and F1, as well as the SAT phase-transition setting. We find that these metrics can be misleading: many models obtain high scores by over-predicting satisfiab"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.28602","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/2605.28602/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-28T02:04:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2clyxUSDK4hWDqdlKo1yEk+2pENnxkVhpCh+nQsO48rWoohpGNxo9yHcInURseDbdDiiId85iANqSdj2OrHIAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-30T17:27:09.432271Z"},"content_sha256":"4cbe599fcc711512802a03668476f34651b6e1e333893b0b95295a483b3e3673","schema_version":"1.0","event_id":"sha256:4cbe599fcc711512802a03668476f34651b6e1e333893b0b95295a483b3e3673"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/DTINKQS5IM2L4MFEVRCWMM2Z3H/bundle.json","state_url":"https://pith.science/pith/DTINKQS5IM2L4MFEVRCWMM2Z3H/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/DTINKQS5IM2L4MFEVRCWMM2Z3H/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-06-30T17:27:09Z","links":{"resolver":"https://pith.science/pith/DTINKQS5IM2L4MFEVRCWMM2Z3H","bundle":"https://pith.science/pith/DTINKQS5IM2L4MFEVRCWMM2Z3H/bundle.json","state":"https://pith.science/pith/DTINKQS5IM2L4MFEVRCWMM2Z3H/state.json","well_known_bundle":"https://pith.science/.well-known/pith/DTINKQS5IM2L4MFEVRCWMM2Z3H/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:DTINKQS5IM2L4MFEVRCWMM2Z3H","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"acf18ca6d3c4aefe1e56451907ee9e9adf9aa89b803975eec9e55ab3d656d5a8","cross_cats_sorted":["cs.CL","cs.LO"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-27T15:18:45Z","title_canon_sha256":"635831c300b28016780559cfd917ac4e49ce998627aad05ae67a6ac4e7860867"},"schema_version":"1.0","source":{"id":"2605.28602","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.28602","created_at":"2026-05-28T02:04:57Z"},{"alias_kind":"arxiv_version","alias_value":"2605.28602v1","created_at":"2026-05-28T02:04:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.28602","created_at":"2026-05-28T02:04:57Z"},{"alias_kind":"pith_short_12","alias_value":"DTINKQS5IM2L","created_at":"2026-05-28T02:04:57Z"},{"alias_kind":"pith_short_16","alias_value":"DTINKQS5IM2L4MFE","created_at":"2026-05-28T02:04:57Z"},{"alias_kind":"pith_short_8","alias_value":"DTINKQS5","created_at":"2026-05-28T02:04:57Z"}],"graph_snapshots":[{"event_id":"sha256:4cbe599fcc711512802a03668476f34651b6e1e333893b0b95295a483b3e3673","target":"graph","created_at":"2026-05-28T02:04:57Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2605.28602/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs) are increasingly used for tasks that implicitly reduce to Boolean satisfiability (SAT), yet their reasoning ability on SAT remains unclear. We present a systematic study of LLMs on 2-SAT and 3-SAT, together with two canonical reductions, Vertex Cover and discrete 3D packing, to probe representation-invariant reasoning. We first evaluate models using conventional metrics, including accuracy, precision, recall, and F1, as well as the SAT phase-transition setting. We find that these metrics can be misleading: many models obtain high scores by over-predicting satisfiab","authors_text":"Leizhen Zhang, Sheng Chen, Shuhan Chen","cross_cats":["cs.CL","cs.LO"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-27T15:18:45Z","title":"Satisfiability Solving with LLMs: A Matched-Pair Evaluation of Reasoning Capability"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.28602","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:83c7f58bfc68ac559be6ba8a7d156390542af09d9b98684d230fd2c8e106244e","target":"record","created_at":"2026-05-28T02:04:57Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"acf18ca6d3c4aefe1e56451907ee9e9adf9aa89b803975eec9e55ab3d656d5a8","cross_cats_sorted":["cs.CL","cs.LO"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-27T15:18:45Z","title_canon_sha256":"635831c300b28016780559cfd917ac4e49ce998627aad05ae67a6ac4e7860867"},"schema_version":"1.0","source":{"id":"2605.28602","kind":"arxiv","version":1}},"canonical_sha256":"1cd0d5425d4334be30a4ac45663359d9de6a574401d032c64049dda8c3bcaec2","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"1cd0d5425d4334be30a4ac45663359d9de6a574401d032c64049dda8c3bcaec2","first_computed_at":"2026-05-28T02:04:57.511289Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-28T02:04:57.511289Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"zEyLMOSpAEArZOs3zfLnU/B1w4KlDXjt+Bh86sWPclHuQSDj1INheb5XLHjfbbAHcsWah6U0TN6C2X0GMF5KCw==","signature_status":"signed_v1","signed_at":"2026-05-28T02:04:57.511732Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.28602","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:83c7f58bfc68ac559be6ba8a7d156390542af09d9b98684d230fd2c8e106244e","sha256:4cbe599fcc711512802a03668476f34651b6e1e333893b0b95295a483b3e3673"],"state_sha256":"197cf2f3288e1f21bf5eeb8a497af32a575df21d8f3095c233ad0084d8c4e8cb"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"T5RkDnUJuEm8pZj3TXonjp1oR5rvemhJIqqlFf8nZcIgMjdERTp50ep3fn432IF5ZF4VZITDtoHqTP1rDsgXBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-30T17:27:09.434271Z","bundle_sha256":"493c9b55d476d46599b8ea7479c05418b3c0caca2247dae661d736f6f481f77e"}}