{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:2AEWBDFU2LVF6J3LVJNYBXCZWE","short_pith_number":"pith:2AEWBDFU","schema_version":"1.0","canonical_sha256":"d009608cb4d2ea5f276baa5b80dc59b1373aa1122ef479626cc49d48b43e1bc3","source":{"kind":"arxiv","id":"2402.10963","version":2},"attestation_state":"computed","paper":{"title":"GLoRe: When, Where, and How to Improve LLM Reasoning via Global and Local Refinements","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Alex Havrilla, Christoforus Nalmpantis, Eric Hambro, Jane Dwivedi-Yu, Maksym Zhuravinskyi, Roberta Raileanu, Sharath Raparthy","submitted_at":"2024-02-13T20:16:29Z","abstract_excerpt":"State-of-the-art language models can exhibit impressive reasoning refinement capabilities on math, science or coding tasks. However, recent work demonstrates that even the best models struggle to identify \\textit{when and where to refine} without access to external feedback. Outcome-based Reward Models (\\textbf{ORMs}), trained to predict correctness of the final answer indicating when to refine, offer one convenient solution for deciding when to refine. Process Based Reward Models (\\textbf{PRMs}), trained to predict correctness of intermediate steps, can then be used to indicate where to refin"},"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":"2402.10963","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-02-13T20:16:29Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"76f0c35d87bf136ce4df450f223ae1d5bc4528f9c91cdacf5c08e8e89a9f600c","abstract_canon_sha256":"2aa83c4541f62db60117e7af9740b61893ba886495e252189cf312708c9cfd20"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:36:22.566864Z","signature_b64":"1fJWm0ZmckPsCqMs3k3OP8xa6JrcMbEHGnkzzSx/hGeUXZ40DOqMsfRS9cFlXB2EzxUxut4+yBWrDExJ9WrVCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d009608cb4d2ea5f276baa5b80dc59b1373aa1122ef479626cc49d48b43e1bc3","last_reissued_at":"2026-07-05T08:36:22.566442Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:36:22.566442Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"GLoRe: When, Where, and How to Improve LLM Reasoning via Global and Local Refinements","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Alex Havrilla, Christoforus Nalmpantis, Eric Hambro, Jane Dwivedi-Yu, Maksym Zhuravinskyi, Roberta Raileanu, Sharath Raparthy","submitted_at":"2024-02-13T20:16:29Z","abstract_excerpt":"State-of-the-art language models can exhibit impressive reasoning refinement capabilities on math, science or coding tasks. However, recent work demonstrates that even the best models struggle to identify \\textit{when and where to refine} without access to external feedback. Outcome-based Reward Models (\\textbf{ORMs}), trained to predict correctness of the final answer indicating when to refine, offer one convenient solution for deciding when to refine. Process Based Reward Models (\\textbf{PRMs}), trained to predict correctness of intermediate steps, can then be used to indicate where to refin"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.10963","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/2402.10963/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":"2402.10963","created_at":"2026-07-05T08:36:22.566500+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.10963v2","created_at":"2026-07-05T08:36:22.566500+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.10963","created_at":"2026-07-05T08:36:22.566500+00:00"},{"alias_kind":"pith_short_12","alias_value":"2AEWBDFU2LVF","created_at":"2026-07-05T08:36:22.566500+00:00"},{"alias_kind":"pith_short_16","alias_value":"2AEWBDFU2LVF6J3L","created_at":"2026-07-05T08:36:22.566500+00:00"},{"alias_kind":"pith_short_8","alias_value":"2AEWBDFU","created_at":"2026-07-05T08:36:22.566500+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.08346","citing_title":"CATPO: Critique-Augmented Tree Policy Optimization","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2409.12917","citing_title":"Training Language Models to Self-Correct via Reinforcement Learning","ref_index":131,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2AEWBDFU2LVF6J3LVJNYBXCZWE","json":"https://pith.science/pith/2AEWBDFU2LVF6J3LVJNYBXCZWE.json","graph_json":"https://pith.science/api/pith-number/2AEWBDFU2LVF6J3LVJNYBXCZWE/graph.json","events_json":"https://pith.science/api/pith-number/2AEWBDFU2LVF6J3LVJNYBXCZWE/events.json","paper":"https://pith.science/paper/2AEWBDFU"},"agent_actions":{"view_html":"https://pith.science/pith/2AEWBDFU2LVF6J3LVJNYBXCZWE","download_json":"https://pith.science/pith/2AEWBDFU2LVF6J3LVJNYBXCZWE.json","view_paper":"https://pith.science/paper/2AEWBDFU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.10963&json=true","fetch_graph":"https://pith.science/api/pith-number/2AEWBDFU2LVF6J3LVJNYBXCZWE/graph.json","fetch_events":"https://pith.science/api/pith-number/2AEWBDFU2LVF6J3LVJNYBXCZWE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2AEWBDFU2LVF6J3LVJNYBXCZWE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2AEWBDFU2LVF6J3LVJNYBXCZWE/action/storage_attestation","attest_author":"https://pith.science/pith/2AEWBDFU2LVF6J3LVJNYBXCZWE/action/author_attestation","sign_citation":"https://pith.science/pith/2AEWBDFU2LVF6J3LVJNYBXCZWE/action/citation_signature","submit_replication":"https://pith.science/pith/2AEWBDFU2LVF6J3LVJNYBXCZWE/action/replication_record"}},"created_at":"2026-07-05T08:36:22.566500+00:00","updated_at":"2026-07-05T08:36:22.566500+00:00"}