{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:WIDACTC3JDVZL2THJNDLZK3Q5Y","short_pith_number":"pith:WIDACTC3","schema_version":"1.0","canonical_sha256":"b206014c5b48eb95ea674b46bcab70ee23c2b0229ce183a84d0eacdd13ab1bc3","source":{"kind":"arxiv","id":"2401.06081","version":2},"attestation_state":"computed","paper":{"title":"Improving Large Language Models via Fine-grained Reinforcement Learning with Minimum Editing Constraint","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Di Zhang, Fuzheng Zhang, Ji-Rong Wen, Junchen Wan, Kun Zhou, Wayne Xin Zhao, Zhipeng Chen","submitted_at":"2024-01-11T17:58:41Z","abstract_excerpt":"Reinforcement learning (RL) has been widely used in training large language models (LLMs) for preventing unexpected outputs, eg reducing harmfulness and errors. However, existing RL methods mostly adopt the instance-level reward, which is unable to provide fine-grained supervision for complex reasoning tasks, and can not focus on the few key tokens that lead to the incorrectness. To address it, we propose a new RL method named RLMEC that incorporates a generative model as the reward model, which is trained by the erroneous solution rewriting task under the minimum editing constraint, and can p"},"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":"2401.06081","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-01-11T17:58:41Z","cross_cats_sorted":[],"title_canon_sha256":"737abe2ed70a2004e8619f153ba0050da0ff68698502747f1af18a53d51d0286","abstract_canon_sha256":"4c3f629591d9520332e47a002f9065c24125a59b89b451ac25a8c801ee7c7293"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:32:34.824088Z","signature_b64":"A9asJVgJorcBzxrUP9jnEtxU0GPQrfeLZfWkZbq6A/TB0/UtYs7pQB5yY9jJChj+WfRev539bxpMqaKlfJsMBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b206014c5b48eb95ea674b46bcab70ee23c2b0229ce183a84d0eacdd13ab1bc3","last_reissued_at":"2026-07-05T08:32:34.823601Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:32:34.823601Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Improving Large Language Models via Fine-grained Reinforcement Learning with Minimum Editing Constraint","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Di Zhang, Fuzheng Zhang, Ji-Rong Wen, Junchen Wan, Kun Zhou, Wayne Xin Zhao, Zhipeng Chen","submitted_at":"2024-01-11T17:58:41Z","abstract_excerpt":"Reinforcement learning (RL) has been widely used in training large language models (LLMs) for preventing unexpected outputs, eg reducing harmfulness and errors. However, existing RL methods mostly adopt the instance-level reward, which is unable to provide fine-grained supervision for complex reasoning tasks, and can not focus on the few key tokens that lead to the incorrectness. To address it, we propose a new RL method named RLMEC that incorporates a generative model as the reward model, which is trained by the erroneous solution rewriting task under the minimum editing constraint, and can p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.06081","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/2401.06081/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":"2401.06081","created_at":"2026-07-05T08:32:34.823660+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.06081v2","created_at":"2026-07-05T08:32:34.823660+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.06081","created_at":"2026-07-05T08:32:34.823660+00:00"},{"alias_kind":"pith_short_12","alias_value":"WIDACTC3JDVZ","created_at":"2026-07-05T08:32:34.823660+00:00"},{"alias_kind":"pith_short_16","alias_value":"WIDACTC3JDVZL2TH","created_at":"2026-07-05T08:32:34.823660+00:00"},{"alias_kind":"pith_short_8","alias_value":"WIDACTC3","created_at":"2026-07-05T08:32:34.823660+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"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":27,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WIDACTC3JDVZL2THJNDLZK3Q5Y","json":"https://pith.science/pith/WIDACTC3JDVZL2THJNDLZK3Q5Y.json","graph_json":"https://pith.science/api/pith-number/WIDACTC3JDVZL2THJNDLZK3Q5Y/graph.json","events_json":"https://pith.science/api/pith-number/WIDACTC3JDVZL2THJNDLZK3Q5Y/events.json","paper":"https://pith.science/paper/WIDACTC3"},"agent_actions":{"view_html":"https://pith.science/pith/WIDACTC3JDVZL2THJNDLZK3Q5Y","download_json":"https://pith.science/pith/WIDACTC3JDVZL2THJNDLZK3Q5Y.json","view_paper":"https://pith.science/paper/WIDACTC3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.06081&json=true","fetch_graph":"https://pith.science/api/pith-number/WIDACTC3JDVZL2THJNDLZK3Q5Y/graph.json","fetch_events":"https://pith.science/api/pith-number/WIDACTC3JDVZL2THJNDLZK3Q5Y/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WIDACTC3JDVZL2THJNDLZK3Q5Y/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WIDACTC3JDVZL2THJNDLZK3Q5Y/action/storage_attestation","attest_author":"https://pith.science/pith/WIDACTC3JDVZL2THJNDLZK3Q5Y/action/author_attestation","sign_citation":"https://pith.science/pith/WIDACTC3JDVZL2THJNDLZK3Q5Y/action/citation_signature","submit_replication":"https://pith.science/pith/WIDACTC3JDVZL2THJNDLZK3Q5Y/action/replication_record"}},"created_at":"2026-07-05T08:32:34.823660+00:00","updated_at":"2026-07-05T08:32:34.823660+00:00"}