{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:GWO37TFUAMZN7MX5BRGNSGYXV7","short_pith_number":"pith:GWO37TFU","schema_version":"1.0","canonical_sha256":"359dbfccb40332dfb2fd0c4cd91b17afdb808013fcf74d5e96d7489f3743adfe","source":{"kind":"arxiv","id":"2607.23731","version":1},"attestation_state":"computed","paper":{"title":"Outcome-Confounded Local Supervision in On-Policy Distillation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Guoqing Ma","submitted_at":"2026-07-26T16:03:33Z","abstract_excerpt":"On-policy distillation (OPD) trains a student on its own trajectories while a teacher supplies dense token-level likelihoods at student-visited prefixes. These likelihoods are often read locally: agreement appears safe to imitate, whereas disagreement appears to identify an error. We show that both readings are confounded by the outcome of the completed trajectory. We introduce an outcome-resolved diagnostic that crosses pointwise teacher-student divergence with final-answer correctness, separating safe imitation, productive divergence, harmful divergence, and agreement-on-failure. In an eight"},"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":"2607.23731","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-07-26T16:03:33Z","cross_cats_sorted":[],"title_canon_sha256":"04aa566e34387fee8a03b104e2d7401f63aa22255721c23f3f69f273c248d824","abstract_canon_sha256":"dcca3ed177b1925f375e08da2a07dafb23fc45176f5e518621bf3562708568af"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-28T01:23:07.929138Z","signature_b64":"KzqBTCM7EXkyod8tojMM91I9N3JOPAKAFCkF2yElREFzpXdPBAnINEG09hUK9J1hKIb2bhCeM8yPcU7VG8ZGAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"359dbfccb40332dfb2fd0c4cd91b17afdb808013fcf74d5e96d7489f3743adfe","last_reissued_at":"2026-07-28T01:23:07.928284Z","signature_status":"signed_v1","first_computed_at":"2026-07-28T01:23:07.928284Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Outcome-Confounded Local Supervision in On-Policy Distillation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Guoqing Ma","submitted_at":"2026-07-26T16:03:33Z","abstract_excerpt":"On-policy distillation (OPD) trains a student on its own trajectories while a teacher supplies dense token-level likelihoods at student-visited prefixes. These likelihoods are often read locally: agreement appears safe to imitate, whereas disagreement appears to identify an error. We show that both readings are confounded by the outcome of the completed trajectory. We introduce an outcome-resolved diagnostic that crosses pointwise teacher-student divergence with final-answer correctness, separating safe imitation, productive divergence, harmful divergence, and agreement-on-failure. In an eight"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.23731","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/2607.23731/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":"2607.23731","created_at":"2026-07-28T01:23:07.928738+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.23731v1","created_at":"2026-07-28T01:23:07.928738+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.23731","created_at":"2026-07-28T01:23:07.928738+00:00"},{"alias_kind":"pith_short_12","alias_value":"GWO37TFUAMZN","created_at":"2026-07-28T01:23:07.928738+00:00"},{"alias_kind":"pith_short_16","alias_value":"GWO37TFUAMZN7MX5","created_at":"2026-07-28T01:23:07.928738+00:00"},{"alias_kind":"pith_short_8","alias_value":"GWO37TFU","created_at":"2026-07-28T01:23:07.928738+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/GWO37TFUAMZN7MX5BRGNSGYXV7","json":"https://pith.science/pith/GWO37TFUAMZN7MX5BRGNSGYXV7.json","graph_json":"https://pith.science/api/pith-number/GWO37TFUAMZN7MX5BRGNSGYXV7/graph.json","events_json":"https://pith.science/api/pith-number/GWO37TFUAMZN7MX5BRGNSGYXV7/events.json","paper":"https://pith.science/paper/GWO37TFU"},"agent_actions":{"view_html":"https://pith.science/pith/GWO37TFUAMZN7MX5BRGNSGYXV7","download_json":"https://pith.science/pith/GWO37TFUAMZN7MX5BRGNSGYXV7.json","view_paper":"https://pith.science/paper/GWO37TFU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.23731&json=true","fetch_graph":"https://pith.science/api/pith-number/GWO37TFUAMZN7MX5BRGNSGYXV7/graph.json","fetch_events":"https://pith.science/api/pith-number/GWO37TFUAMZN7MX5BRGNSGYXV7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GWO37TFUAMZN7MX5BRGNSGYXV7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GWO37TFUAMZN7MX5BRGNSGYXV7/action/storage_attestation","attest_author":"https://pith.science/pith/GWO37TFUAMZN7MX5BRGNSGYXV7/action/author_attestation","sign_citation":"https://pith.science/pith/GWO37TFUAMZN7MX5BRGNSGYXV7/action/citation_signature","submit_replication":"https://pith.science/pith/GWO37TFUAMZN7MX5BRGNSGYXV7/action/replication_record"}},"created_at":"2026-07-28T01:23:07.928738+00:00","updated_at":"2026-07-28T01:23:07.928738+00:00"}