{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:WIDACTC3JDVZL2THJNDLZK3Q5Y","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":"4c3f629591d9520332e47a002f9065c24125a59b89b451ac25a8c801ee7c7293","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-01-11T17:58:41Z","title_canon_sha256":"737abe2ed70a2004e8619f153ba0050da0ff68698502747f1af18a53d51d0286"},"schema_version":"1.0","source":{"id":"2401.06081","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2401.06081","created_at":"2026-07-05T08:32:34Z"},{"alias_kind":"arxiv_version","alias_value":"2401.06081v2","created_at":"2026-07-05T08:32:34Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.06081","created_at":"2026-07-05T08:32:34Z"},{"alias_kind":"pith_short_12","alias_value":"WIDACTC3JDVZ","created_at":"2026-07-05T08:32:34Z"},{"alias_kind":"pith_short_16","alias_value":"WIDACTC3JDVZL2TH","created_at":"2026-07-05T08:32:34Z"},{"alias_kind":"pith_short_8","alias_value":"WIDACTC3","created_at":"2026-07-05T08:32:34Z"}],"graph_snapshots":[{"event_id":"sha256:9496550562ba63aa5f1d3a3f10ebcad6f865323bc473cef396017e4d0d96ce0b","target":"graph","created_at":"2026-07-05T08:32:34Z","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/2401.06081/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"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","authors_text":"Di Zhang, Fuzheng Zhang, Ji-Rong Wen, Junchen Wan, Kun Zhou, Wayne Xin Zhao, Zhipeng Chen","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-01-11T17:58:41Z","title":"Improving Large Language Models via Fine-grained Reinforcement Learning with Minimum Editing Constraint"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.06081","kind":"arxiv","version":2},"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:a3b95d5ad9a3aa99a0427faa15a9c1117bcaaab978c69e2e5cca07d1a59ef16a","target":"record","created_at":"2026-07-05T08:32:34Z","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":"4c3f629591d9520332e47a002f9065c24125a59b89b451ac25a8c801ee7c7293","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-01-11T17:58:41Z","title_canon_sha256":"737abe2ed70a2004e8619f153ba0050da0ff68698502747f1af18a53d51d0286"},"schema_version":"1.0","source":{"id":"2401.06081","kind":"arxiv","version":2}},"canonical_sha256":"b206014c5b48eb95ea674b46bcab70ee23c2b0229ce183a84d0eacdd13ab1bc3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b206014c5b48eb95ea674b46bcab70ee23c2b0229ce183a84d0eacdd13ab1bc3","first_computed_at":"2026-07-05T08:32:34.823601Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:32:34.823601Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"A9asJVgJorcBzxrUP9jnEtxU0GPQrfeLZfWkZbq6A/TB0/UtYs7pQB5yY9jJChj+WfRev539bxpMqaKlfJsMBg==","signature_status":"signed_v1","signed_at":"2026-07-05T08:32:34.824088Z","signed_message":"canonical_sha256_bytes"},"source_id":"2401.06081","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a3b95d5ad9a3aa99a0427faa15a9c1117bcaaab978c69e2e5cca07d1a59ef16a","sha256:9496550562ba63aa5f1d3a3f10ebcad6f865323bc473cef396017e4d0d96ce0b"],"state_sha256":"cd2a4ea4c7f7d3f0efd94096a60007d1d25771c7eb2c2f9920772273b5cbb833"}