{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:MXQTO4WXWZ5WU5MTATLMBHZZYD","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":"10e7d04bf3e837bbb599f565529c91c66acdd6a96e9a88ef4d9d3ec71cea9221","cross_cats_sorted":["cs.CL","cs.LG","cs.LO"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-12-06T10:49:50Z","title_canon_sha256":"174223e13832395cc26d800e70f119f22c2332f40073368839b4628477570eff"},"schema_version":"1.0","source":{"id":"2512.06393","kind":"arxiv","version":7}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2512.06393","created_at":"2026-05-26T01:03:19Z"},{"alias_kind":"arxiv_version","alias_value":"2512.06393v7","created_at":"2026-05-26T01:03:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.06393","created_at":"2026-05-26T01:03:19Z"},{"alias_kind":"pith_short_12","alias_value":"MXQTO4WXWZ5W","created_at":"2026-05-26T01:03:19Z"},{"alias_kind":"pith_short_16","alias_value":"MXQTO4WXWZ5WU5MT","created_at":"2026-05-26T01:03:19Z"},{"alias_kind":"pith_short_8","alias_value":"MXQTO4WX","created_at":"2026-05-26T01:03:19Z"}],"graph_snapshots":[{"event_id":"sha256:a8f83f961244a6e3a45f66b955ae80fa8adaaa48ae95b29d0805727efc86687b","target":"graph","created_at":"2026-05-26T01:03:19Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"The pipeline saturates all four primary stress tests for both 1.5B and 8B backbones. We further validate a Phase 2 extension that replaces the propositional oracle with a Lean 4 kernel, attaining 99.0% kernel agreement on the 105 classically-derivable (T) questions within a stratified 187-question Lean-translated sample (overall 71.7% across both polarities)."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the four stress tests capture the relevant forms of logical inconsistency that arise in real-world LLM use, and that the symbolic forward-chaining engine (or Lean kernel) supplies rewards that do not embed their own unexamined biases or coverage gaps into the learned policy."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Conflict-Aware Fusion mitigates Logic Inertia in LLMs through a four-stage pipeline of SFT, DPO, logical invariance regularization, and reinforcement learning from a symbolic oracle, saturating four stress tests on rule contradictions."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"A four-stage training pipeline makes LLMs check rules for contradictions before reasoning, fixing their tendency to follow inconsistent premises."}],"snapshot_sha256":"3612cf4edbef4804ebd6e2f8040131b98825c0319488755a5287036837593de6"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"2e79b3fce7d1bc5803778319a7763cc2ee386c5580d081416245cd9eec7c3e83"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2512.06393/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs) achieve high accuracy on many reasoning benchmarks but remain brittle under structural perturbations of rule-based systems. We introduce a diagnostic framework with four stress tests -- redundant vs. essential rule deletion, contradictory-rule injection, logic-preserving rewrites, and multi-law stacking -- and use it to expose Logic Inertia: the tendency of generative LLMs (Qwen2/3, TinyLlama, GPT-4o, Gemma-3-4B-IT) and the encoder-only BERT baseline to persist along learned deductive trajectories under inconsistent premises. The collapse is sharp: untreated baseli","authors_text":"Michael Witbrock, Qiming Bao, Xiaoxuan Fu","cross_cats":["cs.CL","cs.LG","cs.LO"],"headline":"A four-stage training pipeline makes LLMs check rules for contradictions before reasoning, fixing their tendency to follow inconsistent premises.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-12-06T10:49:50Z","title":"Conflict-Aware Fusion: Mitigating Logic Inertia in Large Language Models via Structured Cognitive Priors"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2512.06393","kind":"arxiv","version":7},"verdict":{"created_at":"2026-05-17T01:05:09.434154Z","id":"b5e5e1d1-0277-4db1-8f4b-96d3cc140796","model_set":{"reader":"grok-4.3"},"one_line_summary":"Conflict-Aware Fusion mitigates Logic Inertia in LLMs through a four-stage pipeline of SFT, DPO, logical invariance regularization, and reinforcement learning from a symbolic oracle, saturating four stress tests on rule contradictions.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"A four-stage training pipeline makes LLMs check rules for contradictions before reasoning, fixing their tendency to follow inconsistent premises.","strongest_claim":"The pipeline saturates all four primary stress tests for both 1.5B and 8B backbones. We further validate a Phase 2 extension that replaces the propositional oracle with a Lean 4 kernel, attaining 99.0% kernel agreement on the 105 classically-derivable (T) questions within a stratified 187-question Lean-translated sample (overall 71.7% across both polarities).","weakest_assumption":"That the four stress tests capture the relevant forms of logical inconsistency that arise in real-world LLM use, and that the symbolic forward-chaining engine (or Lean kernel) supplies rewards that do not embed their own unexamined biases or coverage gaps into the learned policy."}},"verdict_id":"b5e5e1d1-0277-4db1-8f4b-96d3cc140796"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:44f5aa65b4e2bf259e3d9efde5b5e806b1f3b5389321addb4f1c3c3b87093f16","target":"record","created_at":"2026-05-26T01:03:19Z","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":"10e7d04bf3e837bbb599f565529c91c66acdd6a96e9a88ef4d9d3ec71cea9221","cross_cats_sorted":["cs.CL","cs.LG","cs.LO"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-12-06T10:49:50Z","title_canon_sha256":"174223e13832395cc26d800e70f119f22c2332f40073368839b4628477570eff"},"schema_version":"1.0","source":{"id":"2512.06393","kind":"arxiv","version":7}},"canonical_sha256":"65e13772d7b67b6a759304d6c09f39c0e1459711a404ca9300d750fdfcea6856","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"65e13772d7b67b6a759304d6c09f39c0e1459711a404ca9300d750fdfcea6856","first_computed_at":"2026-05-26T01:03:19.406900Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-26T01:03:19.406900Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"SLweVWFYIP22ZJbD4g/qwVuhgjePwJBQ50cUkTkiy8U4X8j7P6aUirdgKojjit4Yfgu/M0qbVUQxU8hzo79HAA==","signature_status":"signed_v1","signed_at":"2026-05-26T01:03:19.407477Z","signed_message":"canonical_sha256_bytes"},"source_id":"2512.06393","source_kind":"arxiv","source_version":7}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:44f5aa65b4e2bf259e3d9efde5b5e806b1f3b5389321addb4f1c3c3b87093f16","sha256:a8f83f961244a6e3a45f66b955ae80fa8adaaa48ae95b29d0805727efc86687b"],"state_sha256":"088ac3ce3e93684b935144a3d2b73a27fa940fe824b750bb03b61efb779fe62c"}