{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:PDCTJBTC6GDYS74PKQ3BCJJPGR","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":"91adbad2920422e205f37202aebe55cbf7b3d38603738d2a71ec7336af409f1b","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2023-12-14T13:41:54Z","title_canon_sha256":"58c714679bd1e6ffd8b0af13ba869ae97a1daf04659e61eab2fc800ac1e80bd8"},"schema_version":"1.0","source":{"id":"2312.08935","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2312.08935","created_at":"2026-05-17T23:39:21Z"},{"alias_kind":"arxiv_version","alias_value":"2312.08935v3","created_at":"2026-05-17T23:39:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.08935","created_at":"2026-05-17T23:39:21Z"},{"alias_kind":"pith_short_12","alias_value":"PDCTJBTC6GDY","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"PDCTJBTC6GDYS74P","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"PDCTJBTC","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:ca5c12ae674716018ea48522a6684df09f3f557fc058c2ba271534af29c1bc29","target":"graph","created_at":"2026-05-17T23:39:21Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"the step-by-step PPO with Math-Shepherd significantly improves the accuracy of Mistral-7B (77.9%→84.1% on GSM8K and 28.6%→33.0% on MATH). The accuracy can be further enhanced to 89.1% and 43.5% on GSM8K and MATH with the verification of Math-Shepherd."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That automatically constructed process-wise supervision data accurately labels correct versus incorrect reasoning steps without systematic bias or noise from the generation process itself."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Math-Shepherd is an automatically trained process reward model that scores solution steps to verify and reinforce LLMs, lifting Mistral-7B from 77.9% to 89.1% on GSM8K and 28.6% to 43.5% on MATH."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Math-Shepherd trains reward models on auto-generated step labels to verify and reinforce LLM math solutions without human annotations."}],"snapshot_sha256":"31073cbe9ee9b6bb1cad67334a070dc6e908f2f330ea328d8692dcd2220d7195"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"99ac36ec696a4495b7b2913802107dc7d447e534379a98c764abdc37ee66f038"},"paper":{"abstract_excerpt":"In this paper, we present an innovative process-oriented math process reward model called \\textbf{Math-Shepherd}, which assigns a reward score to each step of math problem solutions. The training of Math-Shepherd is achieved using automatically constructed process-wise supervision data, breaking the bottleneck of heavy reliance on manual annotation in existing work. We explore the effectiveness of Math-Shepherd in two scenarios: 1) \\textit{Verification}: Math-Shepherd is utilized for reranking multiple outputs generated by Large Language Models (LLMs); 2) \\textit{Reinforcement Learning}: Math-","authors_text":"Damai Dai, Deli Chen, Lei Li, Peiyi Wang, R.X. Xu, Yifei Li, Y.Wu, Zhifang Sui, Zhihong Shao","cross_cats":["cs.CL","cs.LG"],"headline":"Math-Shepherd trains reward models on auto-generated step labels to verify and reinforce LLM math solutions without human annotations.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2023-12-14T13:41:54Z","title":"Math-Shepherd: Verify and Reinforce LLMs Step-by-step without Human Annotations"},"references":{"count":60,"internal_anchors":19,"resolved_work":60,"sample":[{"cited_arxiv_id":"","doi":"10.18653/v1/2022.emnlp-main.225","is_internal_anchor":false,"ref_index":1,"title":"Red teaming language models with language models","work_id":"664322b7-6ac6-46c5-b1f2-193a778945d2","year":2022},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":10,"title":"International Conference on Machine Learning , pages=","work_id":"1a650b5b-a768-4080-8629-3f3f3fe0d908","year":2023},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":11,"title":"Proceedings of the 29th Symposium on Operating Systems Principles , pages=","work_id":"942205f6-1365-4509-8f46-700be8023817","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":13,"title":"Chi and Quoc V","work_id":"e5146dbc-54b0-40cd-a84c-3d72af59c83f","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":14,"title":"Xuezhi Wang and Jason Wei and Dale Schuurmans and Quoc V. Le and Ed H. Chi and Sharan Narang and Aakanksha Chowdhery and Denny Zhou , title =. The Eleventh International Conference on Learning Represe","work_id":"871bab98-59d0-4eff-8c2d-56c7055fbe92","year":2023}],"snapshot_sha256":"efd9458cd4f2b4aedc77964b90d0c75b1942d7b59014aec8b5e9018a7f2bae28"},"source":{"id":"2312.08935","kind":"arxiv","version":3},"verdict":{"created_at":"2026-05-14T22:29:38.954816Z","id":"d75c5099-d22d-4fbe-aebf-456ab69fe5f1","model_set":{"reader":"grok-4.3"},"one_line_summary":"Math-Shepherd is an automatically trained process reward model that scores solution steps to verify and reinforce LLMs, lifting Mistral-7B from 77.9% to 89.1% on GSM8K and 28.6% to 43.5% on MATH.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Math-Shepherd trains reward models on auto-generated step labels to verify and reinforce LLM math solutions without human annotations.","strongest_claim":"the step-by-step PPO with Math-Shepherd significantly improves the accuracy of Mistral-7B (77.9%→84.1% on GSM8K and 28.6%→33.0% on MATH). The accuracy can be further enhanced to 89.1% and 43.5% on GSM8K and MATH with the verification of Math-Shepherd.","weakest_assumption":"That automatically constructed process-wise supervision data accurately labels correct versus incorrect reasoning steps without systematic bias or noise from the generation process itself."}},"verdict_id":"d75c5099-d22d-4fbe-aebf-456ab69fe5f1"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:9d1a550b33132cd6edbbd361deaab39d80feca6b320e9da121af7c85bb963148","target":"record","created_at":"2026-05-17T23:39:21Z","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":"91adbad2920422e205f37202aebe55cbf7b3d38603738d2a71ec7336af409f1b","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2023-12-14T13:41:54Z","title_canon_sha256":"58c714679bd1e6ffd8b0af13ba869ae97a1daf04659e61eab2fc800ac1e80bd8"},"schema_version":"1.0","source":{"id":"2312.08935","kind":"arxiv","version":3}},"canonical_sha256":"78c5348662f187897f8f543611252f34475cba6704c6af5bff58952f261625f2","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"78c5348662f187897f8f543611252f34475cba6704c6af5bff58952f261625f2","first_computed_at":"2026-05-17T23:39:21.442498Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:39:21.442498Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"NenUiQfXGeuy1TuEPAfCUplsnfNWrWFE2wwte8MoNneg6QVU86bTPhiJBpNbPc8ZvDUyDtWE6CfAVYriQ/5EDg==","signature_status":"signed_v1","signed_at":"2026-05-17T23:39:21.443166Z","signed_message":"canonical_sha256_bytes"},"source_id":"2312.08935","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9d1a550b33132cd6edbbd361deaab39d80feca6b320e9da121af7c85bb963148","sha256:ca5c12ae674716018ea48522a6684df09f3f557fc058c2ba271534af29c1bc29"],"state_sha256":"250005aff7de36e922066d8936662b50baa86e696acad515cf5e206dab4fea56"}