{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:X3EKF6HQGCFIBAZXRQRQREAESV","short_pith_number":"pith:X3EKF6HQ","schema_version":"1.0","canonical_sha256":"bec8a2f8f0308a8083378c230890049546fc9d37deb39e0517001501555e55de","source":{"kind":"arxiv","id":"2502.19613","version":1},"attestation_state":"computed","paper":{"title":"Self-rewarding correction for mathematical reasoning","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Chenlu Ye, Hanning Zhang, Lichang Chen, Nan Jiang, Tong Zhang, Wei Xiong","submitted_at":"2025-02-26T23:01:16Z","abstract_excerpt":"We study self-rewarding reasoning large language models (LLMs), which can simultaneously generate step-by-step reasoning and evaluate the correctness of their outputs during the inference time-without external feedback. This integrated approach allows a single model to independently guide its reasoning process, offering computational advantages for model deployment. We particularly focus on the representative task of self-correction, where models autonomously detect errors in their responses, revise outputs, and decide when to terminate iterative refinement loops. To enable this, we propose a "},"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":"2502.19613","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2025-02-26T23:01:16Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"8a76fc2bef4d7824f7a2bd7ff36d0d461a61c8641d02874a0cd05648f50dbc81","abstract_canon_sha256":"016d00358164e5b3e636b540eaf6ea3c1b806af8712ca48895db8e874ccace61"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:20:49.870489Z","signature_b64":"5igP9KC2oxWVmmhdPbemkYSsX6IVzHlmfDuwGzl15A0/+XfTU1dFu8Hi8u2KryEp205rYnteYV1mUDg6Rar0Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bec8a2f8f0308a8083378c230890049546fc9d37deb39e0517001501555e55de","last_reissued_at":"2026-07-05T10:20:49.869978Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:20:49.869978Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Self-rewarding correction for mathematical reasoning","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Chenlu Ye, Hanning Zhang, Lichang Chen, Nan Jiang, Tong Zhang, Wei Xiong","submitted_at":"2025-02-26T23:01:16Z","abstract_excerpt":"We study self-rewarding reasoning large language models (LLMs), which can simultaneously generate step-by-step reasoning and evaluate the correctness of their outputs during the inference time-without external feedback. This integrated approach allows a single model to independently guide its reasoning process, offering computational advantages for model deployment. We particularly focus on the representative task of self-correction, where models autonomously detect errors in their responses, revise outputs, and decide when to terminate iterative refinement loops. To enable this, we propose a "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.19613","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/2502.19613/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":"2502.19613","created_at":"2026-07-05T10:20:49.870036+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.19613v1","created_at":"2026-07-05T10:20:49.870036+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.19613","created_at":"2026-07-05T10:20:49.870036+00:00"},{"alias_kind":"pith_short_12","alias_value":"X3EKF6HQGCFI","created_at":"2026-07-05T10:20:49.870036+00:00"},{"alias_kind":"pith_short_16","alias_value":"X3EKF6HQGCFIBAZX","created_at":"2026-07-05T10:20:49.870036+00:00"},{"alias_kind":"pith_short_8","alias_value":"X3EKF6HQ","created_at":"2026-07-05T10:20:49.870036+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":13,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21943","citing_title":"Modularized Reinforcement Learning on LLMs: From MDP Creation to Exploration and Learning","ref_index":238,"is_internal_anchor":false},{"citing_arxiv_id":"2606.13316","citing_title":"ReSum: Synergizing LLM Reasoning and Summarization with Reinforcement Learning","ref_index":64,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01249","citing_title":"Trust Region On-Policy Distillation","ref_index":162,"is_internal_anchor":false},{"citing_arxiv_id":"2606.28186","citing_title":"Cognitive Episodes in LLM Reasoning Traces Enable Interpretable Human Item Difficulty Prediction","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.24613","citing_title":"Guarded Repair for Harm-Aware Post-hoc Replacement of LLM Mathematical Reasoning","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11470","citing_title":"The Periodic Table of LLM Reasoning: A Structured Survey of Reasoning Paradigms, Methods, and Failure Modes","ref_index":274,"is_internal_anchor":false},{"citing_arxiv_id":"2510.07962","citing_title":"LightReasoner: Can Small Language Models Teach Large Language Models Reasoning?","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16299","citing_title":"ACE: Self-Evolving LLM Coding Framework via Adversarial Unit Test Generation and Preference Optimization","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2509.03403","citing_title":"Beyond Correctness: Harmonizing Process and Outcome Rewards through RL Training","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2602.07832","citing_title":"rePIRL: Learn PRM with Inverse RL for LLM Reasoning","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16299","citing_title":"ACE: Self-Evolving LLM Coding Framework via Adversarial Unit Test Generation and Preference Optimization","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2604.03993","citing_title":"Can LLMs Learn to Reason Robustly under Noisy Supervision?","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05226","citing_title":"Internalizing Outcome Supervision into Process Supervision: A New Paradigm for Reinforcement Learning for Reasoning","ref_index":24,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/X3EKF6HQGCFIBAZXRQRQREAESV","json":"https://pith.science/pith/X3EKF6HQGCFIBAZXRQRQREAESV.json","graph_json":"https://pith.science/api/pith-number/X3EKF6HQGCFIBAZXRQRQREAESV/graph.json","events_json":"https://pith.science/api/pith-number/X3EKF6HQGCFIBAZXRQRQREAESV/events.json","paper":"https://pith.science/paper/X3EKF6HQ"},"agent_actions":{"view_html":"https://pith.science/pith/X3EKF6HQGCFIBAZXRQRQREAESV","download_json":"https://pith.science/pith/X3EKF6HQGCFIBAZXRQRQREAESV.json","view_paper":"https://pith.science/paper/X3EKF6HQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.19613&json=true","fetch_graph":"https://pith.science/api/pith-number/X3EKF6HQGCFIBAZXRQRQREAESV/graph.json","fetch_events":"https://pith.science/api/pith-number/X3EKF6HQGCFIBAZXRQRQREAESV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/X3EKF6HQGCFIBAZXRQRQREAESV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/X3EKF6HQGCFIBAZXRQRQREAESV/action/storage_attestation","attest_author":"https://pith.science/pith/X3EKF6HQGCFIBAZXRQRQREAESV/action/author_attestation","sign_citation":"https://pith.science/pith/X3EKF6HQGCFIBAZXRQRQREAESV/action/citation_signature","submit_replication":"https://pith.science/pith/X3EKF6HQGCFIBAZXRQRQREAESV/action/replication_record"}},"created_at":"2026-07-05T10:20:49.870036+00:00","updated_at":"2026-07-05T10:20:49.870036+00:00"}