{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:MEOBRDCOHNMIHMB5UQMWDWHXU5","short_pith_number":"pith:MEOBRDCO","schema_version":"1.0","canonical_sha256":"611c188c4e3b5883b03da41961d8f7a753da2e3b9e731e86a65980355575f55c","source":{"kind":"arxiv","id":"2605.29648","version":1},"attestation_state":"computed","paper":{"title":"Verifiable Rewards Beyond Math and Code: Lightweight Corpus-Grounded Process Supervision for Factual Question Answering","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dehai Min, Haochang Hao, Lu Cheng, Philip S. Yu, Shicheng Fan, Weihao Liu","submitted_at":"2026-05-28T09:14:37Z","abstract_excerpt":"Applying reinforcement learning to improve factual accuracy in knowledge-intensive question answering faces a reward design dilemma. Response-level rewards provide only coarse supervision and cannot distinguish correct from incorrect statements within a reasoning trace. Sentence-level alternatives offer finer-grained feedback, but typically rely on NLI verifiers, LLM judges, or knowledge-verification pipelines that are expensive to deploy at RL scale and often unreliable for rare-entity facts, where accurate reward signals are especially important. We propose CorVer (Corpus Verify), a lightwei"},"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":"2605.29648","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-28T09:14:37Z","cross_cats_sorted":[],"title_canon_sha256":"9bb7cd4eb56530875fbabade83c5abbd69ad3383f6deb7ae2905660f0b3b4697","abstract_canon_sha256":"794f4df49d5ff98c886104ef6006ef0bfa403be1a793b11f5fc4bcd45550e142"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-29T01:05:52.967504Z","signature_b64":"xfgovapGXDzSEvJi/ND7W6ffYOX7chBSqf8Uy7YlqvgCGyv+AYWgSTlTa4gnJhAfRbiyya+mAAPsZYHYzFFFCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"611c188c4e3b5883b03da41961d8f7a753da2e3b9e731e86a65980355575f55c","last_reissued_at":"2026-05-29T01:05:52.966871Z","signature_status":"signed_v1","first_computed_at":"2026-05-29T01:05:52.966871Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Verifiable Rewards Beyond Math and Code: Lightweight Corpus-Grounded Process Supervision for Factual Question Answering","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dehai Min, Haochang Hao, Lu Cheng, Philip S. Yu, Shicheng Fan, Weihao Liu","submitted_at":"2026-05-28T09:14:37Z","abstract_excerpt":"Applying reinforcement learning to improve factual accuracy in knowledge-intensive question answering faces a reward design dilemma. Response-level rewards provide only coarse supervision and cannot distinguish correct from incorrect statements within a reasoning trace. Sentence-level alternatives offer finer-grained feedback, but typically rely on NLI verifiers, LLM judges, or knowledge-verification pipelines that are expensive to deploy at RL scale and often unreliable for rare-entity facts, where accurate reward signals are especially important. We propose CorVer (Corpus Verify), a lightwei"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.29648","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/2605.29648/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":"2605.29648","created_at":"2026-05-29T01:05:52.966979+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.29648v1","created_at":"2026-05-29T01:05:52.966979+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.29648","created_at":"2026-05-29T01:05:52.966979+00:00"},{"alias_kind":"pith_short_12","alias_value":"MEOBRDCOHNMI","created_at":"2026-05-29T01:05:52.966979+00:00"},{"alias_kind":"pith_short_16","alias_value":"MEOBRDCOHNMIHMB5","created_at":"2026-05-29T01:05:52.966979+00:00"},{"alias_kind":"pith_short_8","alias_value":"MEOBRDCO","created_at":"2026-05-29T01:05:52.966979+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/MEOBRDCOHNMIHMB5UQMWDWHXU5","json":"https://pith.science/pith/MEOBRDCOHNMIHMB5UQMWDWHXU5.json","graph_json":"https://pith.science/api/pith-number/MEOBRDCOHNMIHMB5UQMWDWHXU5/graph.json","events_json":"https://pith.science/api/pith-number/MEOBRDCOHNMIHMB5UQMWDWHXU5/events.json","paper":"https://pith.science/paper/MEOBRDCO"},"agent_actions":{"view_html":"https://pith.science/pith/MEOBRDCOHNMIHMB5UQMWDWHXU5","download_json":"https://pith.science/pith/MEOBRDCOHNMIHMB5UQMWDWHXU5.json","view_paper":"https://pith.science/paper/MEOBRDCO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.29648&json=true","fetch_graph":"https://pith.science/api/pith-number/MEOBRDCOHNMIHMB5UQMWDWHXU5/graph.json","fetch_events":"https://pith.science/api/pith-number/MEOBRDCOHNMIHMB5UQMWDWHXU5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MEOBRDCOHNMIHMB5UQMWDWHXU5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MEOBRDCOHNMIHMB5UQMWDWHXU5/action/storage_attestation","attest_author":"https://pith.science/pith/MEOBRDCOHNMIHMB5UQMWDWHXU5/action/author_attestation","sign_citation":"https://pith.science/pith/MEOBRDCOHNMIHMB5UQMWDWHXU5/action/citation_signature","submit_replication":"https://pith.science/pith/MEOBRDCOHNMIHMB5UQMWDWHXU5/action/replication_record"}},"created_at":"2026-05-29T01:05:52.966979+00:00","updated_at":"2026-05-29T01:05:52.966979+00:00"}