{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:CH2JSDHCQ6EYWJCFG7DLAGI23Q","short_pith_number":"pith:CH2JSDHC","schema_version":"1.0","canonical_sha256":"11f4990ce287898b244537c6b0191adc3a84b5e437f38d90a218b424b42ca95a","source":{"kind":"arxiv","id":"2505.14674","version":1},"attestation_state":"computed","paper":{"title":"Reward Reasoning Model","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Furu Wei, Jiaxin Guo, Li Dong, Qingxiu Dong, Shaohan Huang, Xun Wu, Zewen Chi","submitted_at":"2025-05-20T17:58:03Z","abstract_excerpt":"Reward models play a critical role in guiding large language models toward outputs that align with human expectations. However, an open challenge remains in effectively utilizing test-time compute to enhance reward model performance. In this work, we introduce Reward Reasoning Models (RRMs), which are specifically designed to execute a deliberate reasoning process before generating final rewards. Through chain-of-thought reasoning, RRMs leverage additional test-time compute for complex queries where appropriate rewards are not immediately apparent. To develop RRMs, we implement a reinforcement"},"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":"2505.14674","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-20T17:58:03Z","cross_cats_sorted":[],"title_canon_sha256":"786b7b2a52651e19959c3fce673ad3352903eb57c351da9ee9d9a6fd814284e2","abstract_canon_sha256":"235c9ffdaeea6dd0ec8599439ad7994c1795b800f6d22b9c18bc49376591c63b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:06:17.454048Z","signature_b64":"BymEHtGK2C41Jd9d1DcL+AAp1ZWOpzeFylETEN0L/xsnT80n2GjraDi3VbP0h/WAGfbwOZW/DodeLhUG+g1iAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"11f4990ce287898b244537c6b0191adc3a84b5e437f38d90a218b424b42ca95a","last_reissued_at":"2026-07-05T11:06:17.453526Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:06:17.453526Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reward Reasoning Model","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Furu Wei, Jiaxin Guo, Li Dong, Qingxiu Dong, Shaohan Huang, Xun Wu, Zewen Chi","submitted_at":"2025-05-20T17:58:03Z","abstract_excerpt":"Reward models play a critical role in guiding large language models toward outputs that align with human expectations. However, an open challenge remains in effectively utilizing test-time compute to enhance reward model performance. In this work, we introduce Reward Reasoning Models (RRMs), which are specifically designed to execute a deliberate reasoning process before generating final rewards. Through chain-of-thought reasoning, RRMs leverage additional test-time compute for complex queries where appropriate rewards are not immediately apparent. To develop RRMs, we implement a reinforcement"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.14674","kind":"arxiv","version":1},"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/2505.14674/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":"2505.14674","created_at":"2026-07-05T11:06:17.453592+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.14674v1","created_at":"2026-07-05T11:06:17.453592+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.14674","created_at":"2026-07-05T11:06:17.453592+00:00"},{"alias_kind":"pith_short_12","alias_value":"CH2JSDHCQ6EY","created_at":"2026-07-05T11:06:17.453592+00:00"},{"alias_kind":"pith_short_16","alias_value":"CH2JSDHCQ6EYWJCF","created_at":"2026-07-05T11:06:17.453592+00:00"},{"alias_kind":"pith_short_8","alias_value":"CH2JSDHC","created_at":"2026-07-05T11:06:17.453592+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":8,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21627","citing_title":"Counsel: A Meta-Evaluation Dataset for Agentic Tasks","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2606.08147","citing_title":"Biological Reasoning-Informed Regression for Interpretable Regulatory DNA Activity Prediction","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01249","citing_title":"Trust Region On-Policy Distillation","ref_index":155,"is_internal_anchor":false},{"citing_arxiv_id":"2508.03556","citing_title":"VRPRM: Process Reward Modeling via Visual Reasoning","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2509.00084","citing_title":"Learning to Refine: Self-Refinement of Parallel Reasoning in LLMs","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2510.24235","citing_title":"PaTaRM: Bridging Pairwise and Pointwise Signals via Preference-Aware Task-Adaptive Reward Modeling","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2509.08827","citing_title":"A Survey of Reinforcement Learning for Large Reasoning Models","ref_index":173,"is_internal_anchor":false},{"citing_arxiv_id":"2507.17746","citing_title":"Rubrics as Rewards: Reinforcement Learning Beyond Verifiable Domains","ref_index":12,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CH2JSDHCQ6EYWJCFG7DLAGI23Q","json":"https://pith.science/pith/CH2JSDHCQ6EYWJCFG7DLAGI23Q.json","graph_json":"https://pith.science/api/pith-number/CH2JSDHCQ6EYWJCFG7DLAGI23Q/graph.json","events_json":"https://pith.science/api/pith-number/CH2JSDHCQ6EYWJCFG7DLAGI23Q/events.json","paper":"https://pith.science/paper/CH2JSDHC"},"agent_actions":{"view_html":"https://pith.science/pith/CH2JSDHCQ6EYWJCFG7DLAGI23Q","download_json":"https://pith.science/pith/CH2JSDHCQ6EYWJCFG7DLAGI23Q.json","view_paper":"https://pith.science/paper/CH2JSDHC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.14674&json=true","fetch_graph":"https://pith.science/api/pith-number/CH2JSDHCQ6EYWJCFG7DLAGI23Q/graph.json","fetch_events":"https://pith.science/api/pith-number/CH2JSDHCQ6EYWJCFG7DLAGI23Q/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CH2JSDHCQ6EYWJCFG7DLAGI23Q/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CH2JSDHCQ6EYWJCFG7DLAGI23Q/action/storage_attestation","attest_author":"https://pith.science/pith/CH2JSDHCQ6EYWJCFG7DLAGI23Q/action/author_attestation","sign_citation":"https://pith.science/pith/CH2JSDHCQ6EYWJCFG7DLAGI23Q/action/citation_signature","submit_replication":"https://pith.science/pith/CH2JSDHCQ6EYWJCFG7DLAGI23Q/action/replication_record"}},"created_at":"2026-07-05T11:06:17.453592+00:00","updated_at":"2026-07-05T11:06:17.453592+00:00"}