{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:2JFOB32WDMLFSUDMS3XM4JPIJ7","short_pith_number":"pith:2JFOB32W","schema_version":"1.0","canonical_sha256":"d24ae0ef561b1659506c96eece25e84ffa5b0efb07531b11fcf85fccbde03e61","source":{"kind":"arxiv","id":"2508.19229","version":2},"attestation_state":"computed","paper":{"title":"StepWiser: Stepwise Generative Judges for Wiser Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Jason Weston, Olga Golovneva, Sainbayar Sukhbaatar, Tong Zhang, Wei Xiong, Weizhe Yuan, Wenting Zhao","submitted_at":"2025-08-26T17:45:05Z","abstract_excerpt":"As models increasingly leverage multi-step reasoning strategies to solve complex problems, supervising the logical validity of these intermediate steps has become a critical research challenge. Process reward models address this by providing step-by-step feedback, but current approaches have two major drawbacks: they typically function as classifiers without providing explanations, and their reliance on supervised fine-tuning with static datasets limits generalization. Inspired by recent advances, we reframe stepwise reward modeling from a classification task to a reasoning task itself. We thu"},"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":"2508.19229","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-08-26T17:45:05Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"313f17978aa04f2b4a84be4ef2c2a503156b070aea7f953f7482456835890521","abstract_canon_sha256":"3480f6b2d3b321caab5dc7e4552d38f6a10c22695a797c0f3bd3245093252176"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:00:10.909256Z","signature_b64":"uQOz8hAyr1Ij1kGoe0r7vXTzC+ntW8+viqyQn5qdO5s0ERwGPjohApxnwda/Jg9wDrXsz5tVHmFmWTfWPSiBAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d24ae0ef561b1659506c96eece25e84ffa5b0efb07531b11fcf85fccbde03e61","last_reissued_at":"2026-07-05T12:00:10.908837Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:00:10.908837Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"StepWiser: Stepwise Generative Judges for Wiser Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Jason Weston, Olga Golovneva, Sainbayar Sukhbaatar, Tong Zhang, Wei Xiong, Weizhe Yuan, Wenting Zhao","submitted_at":"2025-08-26T17:45:05Z","abstract_excerpt":"As models increasingly leverage multi-step reasoning strategies to solve complex problems, supervising the logical validity of these intermediate steps has become a critical research challenge. Process reward models address this by providing step-by-step feedback, but current approaches have two major drawbacks: they typically function as classifiers without providing explanations, and their reliance on supervised fine-tuning with static datasets limits generalization. Inspired by recent advances, we reframe stepwise reward modeling from a classification task to a reasoning task itself. We thu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.19229","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/2508.19229/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":"2508.19229","created_at":"2026-07-05T12:00:10.908893+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.19229v2","created_at":"2026-07-05T12:00:10.908893+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.19229","created_at":"2026-07-05T12:00:10.908893+00:00"},{"alias_kind":"pith_short_12","alias_value":"2JFOB32WDMLF","created_at":"2026-07-05T12:00:10.908893+00:00"},{"alias_kind":"pith_short_16","alias_value":"2JFOB32WDMLFSUDM","created_at":"2026-07-05T12:00:10.908893+00:00"},{"alias_kind":"pith_short_8","alias_value":"2JFOB32W","created_at":"2026-07-05T12:00:10.908893+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2509.03403","citing_title":"Beyond Correctness: Harmonizing Process and Outcome Rewards through RL Training","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20745","citing_title":"The Hidden Signal of Verifier Strictness: Controlling and Improving Step-Wise Verification via Selective Latent Steering","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18799","citing_title":"ReCrit: Transition-Aware Reinforcement Learning for Scientific Critic Reasoning","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15529","citing_title":"Process Rewards with Learned Reliability","ref_index":65,"is_internal_anchor":false},{"citing_arxiv_id":"2509.02547","citing_title":"The Landscape of Agentic Reinforcement Learning for LLMs: A Survey","ref_index":164,"is_internal_anchor":false},{"citing_arxiv_id":"2511.06209","citing_title":"ReProbe: Efficient Test-Time Scaling of Multi-Step Reasoning by Probing Internal States of Large Language Models","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15416","citing_title":"StoSignSGD: Unbiased Structural Stochasticity Fixes SignSGD for Training Large Language Models","ref_index":43,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2JFOB32WDMLFSUDMS3XM4JPIJ7","json":"https://pith.science/pith/2JFOB32WDMLFSUDMS3XM4JPIJ7.json","graph_json":"https://pith.science/api/pith-number/2JFOB32WDMLFSUDMS3XM4JPIJ7/graph.json","events_json":"https://pith.science/api/pith-number/2JFOB32WDMLFSUDMS3XM4JPIJ7/events.json","paper":"https://pith.science/paper/2JFOB32W"},"agent_actions":{"view_html":"https://pith.science/pith/2JFOB32WDMLFSUDMS3XM4JPIJ7","download_json":"https://pith.science/pith/2JFOB32WDMLFSUDMS3XM4JPIJ7.json","view_paper":"https://pith.science/paper/2JFOB32W","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.19229&json=true","fetch_graph":"https://pith.science/api/pith-number/2JFOB32WDMLFSUDMS3XM4JPIJ7/graph.json","fetch_events":"https://pith.science/api/pith-number/2JFOB32WDMLFSUDMS3XM4JPIJ7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2JFOB32WDMLFSUDMS3XM4JPIJ7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2JFOB32WDMLFSUDMS3XM4JPIJ7/action/storage_attestation","attest_author":"https://pith.science/pith/2JFOB32WDMLFSUDMS3XM4JPIJ7/action/author_attestation","sign_citation":"https://pith.science/pith/2JFOB32WDMLFSUDMS3XM4JPIJ7/action/citation_signature","submit_replication":"https://pith.science/pith/2JFOB32WDMLFSUDMS3XM4JPIJ7/action/replication_record"}},"created_at":"2026-07-05T12:00:10.908893+00:00","updated_at":"2026-07-05T12:00:10.908893+00:00"}