{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:VELLFZZDUPFLLWUU3E2HJM7IX2","short_pith_number":"pith:VELLFZZD","schema_version":"1.0","canonical_sha256":"a916b2e723a3cab5da94d93474b3e8bea518cfd8f96377baf134d420c84b80fd","source":{"kind":"arxiv","id":"2602.08499","version":2},"attestation_state":"computed","paper":{"title":"Contextual Rollout Bandits for Reinforcement Learning with Verifiable Rewards","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Deqing Wang, Fuzhen Zhuang, Guojun Yin, Jiajun Chai, Wei Lin, Xiaodong Lu, Xiaohan Wang, Yikun Ban, Yu Luo, Zhijun Chen","submitted_at":"2026-02-09T10:51:58Z","abstract_excerpt":"Reinforcement Learning with Verifiable Rewards (RLVR) is an effective paradigm for improving the reasoning capabilities of large language models. However, existing RLVR methods utilize rollouts in an indiscriminate and short-horizon manner: responses of heterogeneous quality within each prompt are treated uniformly, and historical rollouts are discarded after a single use. This leads to noisy supervision, poor sample efficiency, and suboptimal policy updates. We address these issues by formulating rollout scheduling in RLVR as a contextual bandit problem and proposing a unified neural scheduli"},"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":"2602.08499","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-09T10:51:58Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"f0793e2cebdc249d939d01035265c6c3f89faba999e23dfe2bfd2a808af770c8","abstract_canon_sha256":"5765a44474f9ff641dd167e746f4b73ac763141c7308f6bca44cd56b5ccfb13a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T02:04:05.823072Z","signature_b64":"G6f4RAa12SMMEmbknMEfqlBBVjVIPFa15TzQLAo8/OyEbncPCzBIe0r4vIKjH20DG9eHhASA/Vd917JX0iuOBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a916b2e723a3cab5da94d93474b3e8bea518cfd8f96377baf134d420c84b80fd","last_reissued_at":"2026-05-26T02:04:05.822046Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T02:04:05.822046Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Contextual Rollout Bandits for Reinforcement Learning with Verifiable Rewards","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Deqing Wang, Fuzhen Zhuang, Guojun Yin, Jiajun Chai, Wei Lin, Xiaodong Lu, Xiaohan Wang, Yikun Ban, Yu Luo, Zhijun Chen","submitted_at":"2026-02-09T10:51:58Z","abstract_excerpt":"Reinforcement Learning with Verifiable Rewards (RLVR) is an effective paradigm for improving the reasoning capabilities of large language models. However, existing RLVR methods utilize rollouts in an indiscriminate and short-horizon manner: responses of heterogeneous quality within each prompt are treated uniformly, and historical rollouts are discarded after a single use. This leads to noisy supervision, poor sample efficiency, and suboptimal policy updates. We address these issues by formulating rollout scheduling in RLVR as a contextual bandit problem and proposing a unified neural scheduli"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.08499","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/2602.08499/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":"2602.08499","created_at":"2026-05-26T02:04:05.822200+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.08499v2","created_at":"2026-05-26T02:04:05.822200+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.08499","created_at":"2026-05-26T02:04:05.822200+00:00"},{"alias_kind":"pith_short_12","alias_value":"VELLFZZDUPFL","created_at":"2026-05-26T02:04:05.822200+00:00"},{"alias_kind":"pith_short_16","alias_value":"VELLFZZDUPFLLWUU","created_at":"2026-05-26T02:04:05.822200+00:00"},{"alias_kind":"pith_short_8","alias_value":"VELLFZZD","created_at":"2026-05-26T02:04:05.822200+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2604.00860","citing_title":"Policy Improvement Reinforcement Learning","ref_index":33,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VELLFZZDUPFLLWUU3E2HJM7IX2","json":"https://pith.science/pith/VELLFZZDUPFLLWUU3E2HJM7IX2.json","graph_json":"https://pith.science/api/pith-number/VELLFZZDUPFLLWUU3E2HJM7IX2/graph.json","events_json":"https://pith.science/api/pith-number/VELLFZZDUPFLLWUU3E2HJM7IX2/events.json","paper":"https://pith.science/paper/VELLFZZD"},"agent_actions":{"view_html":"https://pith.science/pith/VELLFZZDUPFLLWUU3E2HJM7IX2","download_json":"https://pith.science/pith/VELLFZZDUPFLLWUU3E2HJM7IX2.json","view_paper":"https://pith.science/paper/VELLFZZD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.08499&json=true","fetch_graph":"https://pith.science/api/pith-number/VELLFZZDUPFLLWUU3E2HJM7IX2/graph.json","fetch_events":"https://pith.science/api/pith-number/VELLFZZDUPFLLWUU3E2HJM7IX2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VELLFZZDUPFLLWUU3E2HJM7IX2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VELLFZZDUPFLLWUU3E2HJM7IX2/action/storage_attestation","attest_author":"https://pith.science/pith/VELLFZZDUPFLLWUU3E2HJM7IX2/action/author_attestation","sign_citation":"https://pith.science/pith/VELLFZZDUPFLLWUU3E2HJM7IX2/action/citation_signature","submit_replication":"https://pith.science/pith/VELLFZZDUPFLLWUU3E2HJM7IX2/action/replication_record"}},"created_at":"2026-05-26T02:04:05.822200+00:00","updated_at":"2026-05-26T02:04:05.822200+00:00"}