{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:EMBLF33J2DQMOBCEPLZ2LHVL2U","short_pith_number":"pith:EMBLF33J","schema_version":"1.0","canonical_sha256":"2302b2ef69d0e0c704447af3a59eabd528ef56afd23e398f107da270857b35c1","source":{"kind":"arxiv","id":"2605.25604","version":1},"attestation_state":"computed","paper":{"title":"DVAO: Dynamic Variance-adaptive Advantage Optimization for Multi-reward Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Chuzhan Hao, Guochao Jiang, Guofeng Quan, Guohua Liu, Jingyi Song, Yuewei Zhang","submitted_at":"2026-05-25T08:55:16Z","abstract_excerpt":"Reinforcement Learning has become a standard paradigm for aligning Large Language Models with human intent and task requirements. While Group Relative Policy Optimization offers an efficient, value-model-free alternative to Proximal Policy Optimization, adapting it to real-world multi-reward settings remains challenging. Standard scalarization practices, such as Reward Combination and Advantage Combination, suffer from significant drawbacks: Reward Combination frequently generates advantages with excessively large squared magnitudes that lead to training instability, while Advantage Combinatio"},"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":"2605.25604","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-25T08:55:16Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"f68e9f0dd4ae9d9b7c41f6395a228f2b09967f15c28550c832b4d8b71c4cf039","abstract_canon_sha256":"e61346bf055a3e243b5ae3d5065f4a01a8d79cafc555bf132896681fe0ffabee"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T02:04:46.103525Z","signature_b64":"QmP0PptmxuBcgJFzdh4WYlGUlo3TxfbtP8CboH6qLsmc+muQWJZuMXOTx+0lqJI26bMHvb2tn0uMe/wlsjk2AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2302b2ef69d0e0c704447af3a59eabd528ef56afd23e398f107da270857b35c1","last_reissued_at":"2026-05-26T02:04:46.102837Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T02:04:46.102837Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DVAO: Dynamic Variance-adaptive Advantage Optimization for Multi-reward Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Chuzhan Hao, Guochao Jiang, Guofeng Quan, Guohua Liu, Jingyi Song, Yuewei Zhang","submitted_at":"2026-05-25T08:55:16Z","abstract_excerpt":"Reinforcement Learning has become a standard paradigm for aligning Large Language Models with human intent and task requirements. While Group Relative Policy Optimization offers an efficient, value-model-free alternative to Proximal Policy Optimization, adapting it to real-world multi-reward settings remains challenging. Standard scalarization practices, such as Reward Combination and Advantage Combination, suffer from significant drawbacks: Reward Combination frequently generates advantages with excessively large squared magnitudes that lead to training instability, while Advantage Combinatio"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.25604","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/2605.25604/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":"2605.25604","created_at":"2026-05-26T02:04:46.102946+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.25604v1","created_at":"2026-05-26T02:04:46.102946+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.25604","created_at":"2026-05-26T02:04:46.102946+00:00"},{"alias_kind":"pith_short_12","alias_value":"EMBLF33J2DQM","created_at":"2026-05-26T02:04:46.102946+00:00"},{"alias_kind":"pith_short_16","alias_value":"EMBLF33J2DQMOBCE","created_at":"2026-05-26T02:04:46.102946+00:00"},{"alias_kind":"pith_short_8","alias_value":"EMBLF33J","created_at":"2026-05-26T02:04:46.102946+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EMBLF33J2DQMOBCEPLZ2LHVL2U","json":"https://pith.science/pith/EMBLF33J2DQMOBCEPLZ2LHVL2U.json","graph_json":"https://pith.science/api/pith-number/EMBLF33J2DQMOBCEPLZ2LHVL2U/graph.json","events_json":"https://pith.science/api/pith-number/EMBLF33J2DQMOBCEPLZ2LHVL2U/events.json","paper":"https://pith.science/paper/EMBLF33J"},"agent_actions":{"view_html":"https://pith.science/pith/EMBLF33J2DQMOBCEPLZ2LHVL2U","download_json":"https://pith.science/pith/EMBLF33J2DQMOBCEPLZ2LHVL2U.json","view_paper":"https://pith.science/paper/EMBLF33J","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.25604&json=true","fetch_graph":"https://pith.science/api/pith-number/EMBLF33J2DQMOBCEPLZ2LHVL2U/graph.json","fetch_events":"https://pith.science/api/pith-number/EMBLF33J2DQMOBCEPLZ2LHVL2U/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EMBLF33J2DQMOBCEPLZ2LHVL2U/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EMBLF33J2DQMOBCEPLZ2LHVL2U/action/storage_attestation","attest_author":"https://pith.science/pith/EMBLF33J2DQMOBCEPLZ2LHVL2U/action/author_attestation","sign_citation":"https://pith.science/pith/EMBLF33J2DQMOBCEPLZ2LHVL2U/action/citation_signature","submit_replication":"https://pith.science/pith/EMBLF33J2DQMOBCEPLZ2LHVL2U/action/replication_record"}},"created_at":"2026-05-26T02:04:46.102946+00:00","updated_at":"2026-05-26T02:04:46.102946+00:00"}