{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:DT7QFW6KCJEFBNM5HIB4ZYXTZC","short_pith_number":"pith:DT7QFW6K","schema_version":"1.0","canonical_sha256":"1cff02dbca124850b59d3a03cce2f3c89e0accc60a53db2f998f0d051f078204","source":{"kind":"arxiv","id":"2607.13501","version":1},"attestation_state":"computed","paper":{"title":"LAPO: Leave-One-Turn Attribution for Self-Generated Process Rewards in Multi-Turn Search Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Jiajun Wu, Qiang Zhu","submitted_at":"2026-07-15T06:55:28Z","abstract_excerpt":"Reinforcement learning for multi-turn search reasoning typically relies on terminal outcome rewards, which cannot distinguish useful, redundant, and harmful intermediate interactions. We propose LAPO, a self-generated process-supervision method based on backward leave-one-turn attribution. For each search turn, LAPO replaces the turn and its retrieval observation with a fixed [DELETE] placeholder and measures the resulting change in the current policy's mean log-likelihood of the gold answer. This Answer-Likelihood Gain estimates the turn's contribution while preserving all downstream interact"},"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":"2607.13501","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-07-15T06:55:28Z","cross_cats_sorted":[],"title_canon_sha256":"0d8777635499cd895f23946ad297289d4f43920e98e67862a8493fb6dfa6cd11","abstract_canon_sha256":"1f156c5e560e272e3a9e02a044f9e715b6bf082b592381974d8dffec4e7f0d34"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-16T01:22:16.183974Z","signature_b64":"399iSs8XkM/EKoZXwxY7ya9FY3iIxM/pXej5aT/3Elu7VlA+a/KBUQL58Dw6YTHmudLSM/wMhEBGI+tPAwzKAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1cff02dbca124850b59d3a03cce2f3c89e0accc60a53db2f998f0d051f078204","last_reissued_at":"2026-07-16T01:22:16.183129Z","signature_status":"signed_v1","first_computed_at":"2026-07-16T01:22:16.183129Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LAPO: Leave-One-Turn Attribution for Self-Generated Process Rewards in Multi-Turn Search Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Jiajun Wu, Qiang Zhu","submitted_at":"2026-07-15T06:55:28Z","abstract_excerpt":"Reinforcement learning for multi-turn search reasoning typically relies on terminal outcome rewards, which cannot distinguish useful, redundant, and harmful intermediate interactions. We propose LAPO, a self-generated process-supervision method based on backward leave-one-turn attribution. For each search turn, LAPO replaces the turn and its retrieval observation with a fixed [DELETE] placeholder and measures the resulting change in the current policy's mean log-likelihood of the gold answer. This Answer-Likelihood Gain estimates the turn's contribution while preserving all downstream interact"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.13501","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/2607.13501/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":"2607.13501","created_at":"2026-07-16T01:22:16.183572+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.13501v1","created_at":"2026-07-16T01:22:16.183572+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.13501","created_at":"2026-07-16T01:22:16.183572+00:00"},{"alias_kind":"pith_short_12","alias_value":"DT7QFW6KCJEF","created_at":"2026-07-16T01:22:16.183572+00:00"},{"alias_kind":"pith_short_16","alias_value":"DT7QFW6KCJEFBNM5","created_at":"2026-07-16T01:22:16.183572+00:00"},{"alias_kind":"pith_short_8","alias_value":"DT7QFW6K","created_at":"2026-07-16T01:22:16.183572+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/DT7QFW6KCJEFBNM5HIB4ZYXTZC","json":"https://pith.science/pith/DT7QFW6KCJEFBNM5HIB4ZYXTZC.json","graph_json":"https://pith.science/api/pith-number/DT7QFW6KCJEFBNM5HIB4ZYXTZC/graph.json","events_json":"https://pith.science/api/pith-number/DT7QFW6KCJEFBNM5HIB4ZYXTZC/events.json","paper":"https://pith.science/paper/DT7QFW6K"},"agent_actions":{"view_html":"https://pith.science/pith/DT7QFW6KCJEFBNM5HIB4ZYXTZC","download_json":"https://pith.science/pith/DT7QFW6KCJEFBNM5HIB4ZYXTZC.json","view_paper":"https://pith.science/paper/DT7QFW6K","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.13501&json=true","fetch_graph":"https://pith.science/api/pith-number/DT7QFW6KCJEFBNM5HIB4ZYXTZC/graph.json","fetch_events":"https://pith.science/api/pith-number/DT7QFW6KCJEFBNM5HIB4ZYXTZC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DT7QFW6KCJEFBNM5HIB4ZYXTZC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DT7QFW6KCJEFBNM5HIB4ZYXTZC/action/storage_attestation","attest_author":"https://pith.science/pith/DT7QFW6KCJEFBNM5HIB4ZYXTZC/action/author_attestation","sign_citation":"https://pith.science/pith/DT7QFW6KCJEFBNM5HIB4ZYXTZC/action/citation_signature","submit_replication":"https://pith.science/pith/DT7QFW6KCJEFBNM5HIB4ZYXTZC/action/replication_record"}},"created_at":"2026-07-16T01:22:16.183572+00:00","updated_at":"2026-07-16T01:22:16.183572+00:00"}