{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:GFZXOLDPMTM5K66QYH7ZODT355","short_pith_number":"pith:GFZXOLDP","schema_version":"1.0","canonical_sha256":"3173772c6f64d9d57bd0c1ff970e7bef721a95c85831f9e21f076b3d86b3744c","source":{"kind":"arxiv","id":"2510.00915","version":4},"attestation_state":"computed","paper":{"title":"Reinforcement Learning with Verifiable yet Noisy Rewards under Imperfect Verifiers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Feng Liu, Gang Niu, Masashi Sugiyama, Tongliang Liu, Wei Wang, Xin-Qiang Cai","submitted_at":"2025-10-01T13:56:44Z","abstract_excerpt":"Reinforcement Learning with Verifiable Rewards (RLVR) replaces costly human labeling with automated verifiers. To reduce verifier hacking, many RLVR systems binarize rewards to $\\{0,1\\}$, but imperfect verifiers inevitably introduce \\emph{false negatives} (rejecting correct answers) and \\emph{false positives} (accepting incorrect ones). We formalize verifier unreliability as a stochastic reward channel with asymmetric noise rates $\\rho_0$ and $\\rho_1$ -- the FP rate and the FN rate, respectively. From this abstraction we derive two lightweight corrections: (i) a \\emph{backward} correction that"},"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":"2510.00915","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-10-01T13:56:44Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"50b14707283d27f30d99764afcacfd6216452044f84fffe573ff13caf6e51258","abstract_canon_sha256":"5851894f5dbd5f5170ad7c783451938a566e9ee2542bd22be737bec05ffe1516"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-25T02:01:07.699956Z","signature_b64":"MrKaImT3C48v2U1ia59M4nsizwLm5uboRxC/LGoF1x6LvRiTtJ3KK2451UgLAD5s5WilvvGKi6opPxTVe35bBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3173772c6f64d9d57bd0c1ff970e7bef721a95c85831f9e21f076b3d86b3744c","last_reissued_at":"2026-05-25T02:01:07.698940Z","signature_status":"signed_v1","first_computed_at":"2026-05-25T02:01:07.698940Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reinforcement Learning with Verifiable yet Noisy Rewards under Imperfect Verifiers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Feng Liu, Gang Niu, Masashi Sugiyama, Tongliang Liu, Wei Wang, Xin-Qiang Cai","submitted_at":"2025-10-01T13:56:44Z","abstract_excerpt":"Reinforcement Learning with Verifiable Rewards (RLVR) replaces costly human labeling with automated verifiers. To reduce verifier hacking, many RLVR systems binarize rewards to $\\{0,1\\}$, but imperfect verifiers inevitably introduce \\emph{false negatives} (rejecting correct answers) and \\emph{false positives} (accepting incorrect ones). We formalize verifier unreliability as a stochastic reward channel with asymmetric noise rates $\\rho_0$ and $\\rho_1$ -- the FP rate and the FN rate, respectively. From this abstraction we derive two lightweight corrections: (i) a \\emph{backward} correction that"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.00915","kind":"arxiv","version":4},"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/2510.00915/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":"2510.00915","created_at":"2026-05-25T02:01:07.699082+00:00"},{"alias_kind":"arxiv_version","alias_value":"2510.00915v4","created_at":"2026-05-25T02:01:07.699082+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.00915","created_at":"2026-05-25T02:01:07.699082+00:00"},{"alias_kind":"pith_short_12","alias_value":"GFZXOLDPMTM5","created_at":"2026-05-25T02:01:07.699082+00:00"},{"alias_kind":"pith_short_16","alias_value":"GFZXOLDPMTM5K66Q","created_at":"2026-05-25T02:01:07.699082+00:00"},{"alias_kind":"pith_short_8","alias_value":"GFZXOLDP","created_at":"2026-05-25T02:01:07.699082+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":8,"internal_anchor_count":8,"sample":[{"citing_arxiv_id":"2602.12579","citing_title":"VI-CuRL: Stabilizing Verifier-Independent RL Reasoning via Confidence-Guided Variance Reduction","ref_index":2,"is_internal_anchor":true},{"citing_arxiv_id":"2605.20164","citing_title":"Not Every Rubric Teaches Equally: Policy-Aware Rubric Rewards for RLVR","ref_index":5,"is_internal_anchor":true},{"citing_arxiv_id":"2602.03452","citing_title":"Beyond Variance: Prompt-Efficient RLVR via Rare-Event Amplification and Bidirectional Pairing","ref_index":2,"is_internal_anchor":true},{"citing_arxiv_id":"2605.06732","citing_title":"On Training in Imagination","ref_index":3,"is_internal_anchor":true},{"citing_arxiv_id":"2604.27358","citing_title":"Safe Bilevel Delegation (SBD): A Formal Framework for Runtime Delegation Safety in Multi-Agent Systems","ref_index":2,"is_internal_anchor":true},{"citing_arxiv_id":"2605.05076","citing_title":"High-Dimensional Statistics: Reflections on Progress and Open Problems","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2605.06732","citing_title":"On Training in Imagination","ref_index":35,"is_internal_anchor":true},{"citing_arxiv_id":"2605.02909","citing_title":"Delay, Plateau, or Collapse: Evaluating the Impact of Systematic Verification Error on RLVR","ref_index":6,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GFZXOLDPMTM5K66QYH7ZODT355","json":"https://pith.science/pith/GFZXOLDPMTM5K66QYH7ZODT355.json","graph_json":"https://pith.science/api/pith-number/GFZXOLDPMTM5K66QYH7ZODT355/graph.json","events_json":"https://pith.science/api/pith-number/GFZXOLDPMTM5K66QYH7ZODT355/events.json","paper":"https://pith.science/paper/GFZXOLDP"},"agent_actions":{"view_html":"https://pith.science/pith/GFZXOLDPMTM5K66QYH7ZODT355","download_json":"https://pith.science/pith/GFZXOLDPMTM5K66QYH7ZODT355.json","view_paper":"https://pith.science/paper/GFZXOLDP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2510.00915&json=true","fetch_graph":"https://pith.science/api/pith-number/GFZXOLDPMTM5K66QYH7ZODT355/graph.json","fetch_events":"https://pith.science/api/pith-number/GFZXOLDPMTM5K66QYH7ZODT355/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GFZXOLDPMTM5K66QYH7ZODT355/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GFZXOLDPMTM5K66QYH7ZODT355/action/storage_attestation","attest_author":"https://pith.science/pith/GFZXOLDPMTM5K66QYH7ZODT355/action/author_attestation","sign_citation":"https://pith.science/pith/GFZXOLDPMTM5K66QYH7ZODT355/action/citation_signature","submit_replication":"https://pith.science/pith/GFZXOLDPMTM5K66QYH7ZODT355/action/replication_record"}},"created_at":"2026-05-25T02:01:07.699082+00:00","updated_at":"2026-05-25T02:01:07.699082+00:00"}