{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:2CRNFYLZXWMN7MNXQCJ5HQYKB4","short_pith_number":"pith:2CRNFYLZ","schema_version":"1.0","canonical_sha256":"d0a2d2e179bd98dfb1b78093d3c30a0f2dc406f25485527e18c22e526496dbe2","source":{"kind":"arxiv","id":"2608.01667","version":1},"attestation_state":"computed","paper":{"title":"TCPO: Turn-Level Credit Policy Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Sicong Liao, Yaohua Tang, Zhi Chen","submitted_at":"2026-08-03T04:01:36Z","abstract_excerpt":"Verifier-guided reinforcement learning has become a powerful paradigm for improving LLM reasoning. In multi-turn settings, models receive a verifier score after each turn and iteratively refine their outputs. Although such scores provide dense feedback, they do not directly provide dense credit: a score measures the quality of the current output, while credit should measure how the current turn changes the refinement trajectory. We propose TCPO, a turn-level credit assignment method for verifier-guided multi-turn RL. TCPO casts credit assignment as score-to-credit conversion and constructs tur"},"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":"2608.01667","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-08-03T04:01:36Z","cross_cats_sorted":[],"title_canon_sha256":"8fbe3cf741cd977cfb0297d79216881df6a4483dafa8f7af6c17d24a261c7696","abstract_canon_sha256":"975660d157ca21bd779636db0248c8e38b498b287e89447b289dfc468f491d55"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-04T02:06:22.751355Z","signature_b64":"q5ake8ti3DF5X3ZVBJslZjtuyAiVCGWO2B56N0lFyLgMuiVzrqTF0YTs0z0Ijlb6/S3t+NlQZnE5vpMbCnZnDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d0a2d2e179bd98dfb1b78093d3c30a0f2dc406f25485527e18c22e526496dbe2","last_reissued_at":"2026-08-04T02:06:22.745442Z","signature_status":"signed_v1","first_computed_at":"2026-08-04T02:06:22.745442Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TCPO: Turn-Level Credit Policy Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Sicong Liao, Yaohua Tang, Zhi Chen","submitted_at":"2026-08-03T04:01:36Z","abstract_excerpt":"Verifier-guided reinforcement learning has become a powerful paradigm for improving LLM reasoning. In multi-turn settings, models receive a verifier score after each turn and iteratively refine their outputs. Although such scores provide dense feedback, they do not directly provide dense credit: a score measures the quality of the current output, while credit should measure how the current turn changes the refinement trajectory. We propose TCPO, a turn-level credit assignment method for verifier-guided multi-turn RL. TCPO casts credit assignment as score-to-credit conversion and constructs tur"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.01667","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/2608.01667/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":"2608.01667","created_at":"2026-08-04T02:06:22.751032+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.01667v1","created_at":"2026-08-04T02:06:22.751032+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.01667","created_at":"2026-08-04T02:06:22.751032+00:00"},{"alias_kind":"pith_short_12","alias_value":"2CRNFYLZXWMN","created_at":"2026-08-04T02:06:22.751032+00:00"},{"alias_kind":"pith_short_16","alias_value":"2CRNFYLZXWMN7MNX","created_at":"2026-08-04T02:06:22.751032+00:00"},{"alias_kind":"pith_short_8","alias_value":"2CRNFYLZ","created_at":"2026-08-04T02:06:22.751032+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/2CRNFYLZXWMN7MNXQCJ5HQYKB4","json":"https://pith.science/pith/2CRNFYLZXWMN7MNXQCJ5HQYKB4.json","graph_json":"https://pith.science/api/pith-number/2CRNFYLZXWMN7MNXQCJ5HQYKB4/graph.json","events_json":"https://pith.science/api/pith-number/2CRNFYLZXWMN7MNXQCJ5HQYKB4/events.json","paper":"https://pith.science/paper/2CRNFYLZ"},"agent_actions":{"view_html":"https://pith.science/pith/2CRNFYLZXWMN7MNXQCJ5HQYKB4","download_json":"https://pith.science/pith/2CRNFYLZXWMN7MNXQCJ5HQYKB4.json","view_paper":"https://pith.science/paper/2CRNFYLZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.01667&json=true","fetch_graph":"https://pith.science/api/pith-number/2CRNFYLZXWMN7MNXQCJ5HQYKB4/graph.json","fetch_events":"https://pith.science/api/pith-number/2CRNFYLZXWMN7MNXQCJ5HQYKB4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2CRNFYLZXWMN7MNXQCJ5HQYKB4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2CRNFYLZXWMN7MNXQCJ5HQYKB4/action/storage_attestation","attest_author":"https://pith.science/pith/2CRNFYLZXWMN7MNXQCJ5HQYKB4/action/author_attestation","sign_citation":"https://pith.science/pith/2CRNFYLZXWMN7MNXQCJ5HQYKB4/action/citation_signature","submit_replication":"https://pith.science/pith/2CRNFYLZXWMN7MNXQCJ5HQYKB4/action/replication_record"}},"created_at":"2026-08-04T02:06:22.751032+00:00","updated_at":"2026-08-04T02:06:22.751032+00:00"}