{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:4XZHADLHCZHRBQGCEELCOGFH23","short_pith_number":"pith:4XZHADLH","schema_version":"1.0","canonical_sha256":"e5f2700d67164f10c0c221162718a7d6fc7c65da9349ff62422a68d178183344","source":{"kind":"arxiv","id":"2605.24005","version":1},"attestation_state":"computed","paper":{"title":"LC-ERD: Mining Latent Logic for Self-Evolving Reasoning via Consistency-Regulated Reward Decomposition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Dianzhi Yu, Irwin King, Jiahong Liu, Jiaming Han, Jinhu Qi, Jiyue Jiang, Xiao Guo, Yanyu Chen, Yifei Zhang, Yu Li, Zheng Wu","submitted_at":"2026-05-19T07:27:50Z","abstract_excerpt":"The evolution of Large Language Model (LLM) reasoning is bottlenecked by the scarcity of high-quality process data. While self-alignment via endogenous rewards offers a solution, mining valid supervision faces three challenges: (1) Label Noise via Mimetic Bias, where rewards prioritize statistical likelihood over logical truth, creating a \"correctness illusion\" that masks compounding errors; (2) Coarse-Grained Supervision, where sparse global outcomes (e.g., in GRPO) fail to provide granular guidance, treating reasoning chains as monolithic; and (3) Distributional Collapse, where signals fail "},"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":"2605.24005","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-19T07:27:50Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"d38da409268f380d13ea2d02a7c77625c0730876f30ab5b1fd157b56a837d297","abstract_canon_sha256":"8bb90e0aea19b5a62ce370c657e0a66bed4a3504de979449eb1a06d9e361b8f0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T01:02:40.635987Z","signature_b64":"TUhyzjD1fsr1Lldm/IdwmmsE6WaMIOVQxt/kRbA6TaRS3uKjfKOCRXKYydyEGi11EzPMHG/1YWqPz9gU1f3NDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e5f2700d67164f10c0c221162718a7d6fc7c65da9349ff62422a68d178183344","last_reissued_at":"2026-05-26T01:02:40.635110Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T01:02:40.635110Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LC-ERD: Mining Latent Logic for Self-Evolving Reasoning via Consistency-Regulated Reward Decomposition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Dianzhi Yu, Irwin King, Jiahong Liu, Jiaming Han, Jinhu Qi, Jiyue Jiang, Xiao Guo, Yanyu Chen, Yifei Zhang, Yu Li, Zheng Wu","submitted_at":"2026-05-19T07:27:50Z","abstract_excerpt":"The evolution of Large Language Model (LLM) reasoning is bottlenecked by the scarcity of high-quality process data. While self-alignment via endogenous rewards offers a solution, mining valid supervision faces three challenges: (1) Label Noise via Mimetic Bias, where rewards prioritize statistical likelihood over logical truth, creating a \"correctness illusion\" that masks compounding errors; (2) Coarse-Grained Supervision, where sparse global outcomes (e.g., in GRPO) fail to provide granular guidance, treating reasoning chains as monolithic; and (3) Distributional Collapse, where signals fail "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.24005","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/2605.24005/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":"2605.24005","created_at":"2026-05-26T01:02:40.635244+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.24005v1","created_at":"2026-05-26T01:02:40.635244+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.24005","created_at":"2026-05-26T01:02:40.635244+00:00"},{"alias_kind":"pith_short_12","alias_value":"4XZHADLHCZHR","created_at":"2026-05-26T01:02:40.635244+00:00"},{"alias_kind":"pith_short_16","alias_value":"4XZHADLHCZHRBQGC","created_at":"2026-05-26T01:02:40.635244+00:00"},{"alias_kind":"pith_short_8","alias_value":"4XZHADLH","created_at":"2026-05-26T01:02:40.635244+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4XZHADLHCZHRBQGCEELCOGFH23","json":"https://pith.science/pith/4XZHADLHCZHRBQGCEELCOGFH23.json","graph_json":"https://pith.science/api/pith-number/4XZHADLHCZHRBQGCEELCOGFH23/graph.json","events_json":"https://pith.science/api/pith-number/4XZHADLHCZHRBQGCEELCOGFH23/events.json","paper":"https://pith.science/paper/4XZHADLH"},"agent_actions":{"view_html":"https://pith.science/pith/4XZHADLHCZHRBQGCEELCOGFH23","download_json":"https://pith.science/pith/4XZHADLHCZHRBQGCEELCOGFH23.json","view_paper":"https://pith.science/paper/4XZHADLH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.24005&json=true","fetch_graph":"https://pith.science/api/pith-number/4XZHADLHCZHRBQGCEELCOGFH23/graph.json","fetch_events":"https://pith.science/api/pith-number/4XZHADLHCZHRBQGCEELCOGFH23/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4XZHADLHCZHRBQGCEELCOGFH23/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4XZHADLHCZHRBQGCEELCOGFH23/action/storage_attestation","attest_author":"https://pith.science/pith/4XZHADLHCZHRBQGCEELCOGFH23/action/author_attestation","sign_citation":"https://pith.science/pith/4XZHADLHCZHRBQGCEELCOGFH23/action/citation_signature","submit_replication":"https://pith.science/pith/4XZHADLHCZHRBQGCEELCOGFH23/action/replication_record"}},"created_at":"2026-05-26T01:02:40.635244+00:00","updated_at":"2026-05-26T01:02:40.635244+00:00"}