{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:7HG5I3ZN5UBBZEUGEC6F3WJ7DF","short_pith_number":"pith:7HG5I3ZN","schema_version":"1.0","canonical_sha256":"f9cdd46f2ded021c928620bc5dd93f1960db15b8557032eb8cb6bad172965fbd","source":{"kind":"arxiv","id":"2605.17497","version":1},"attestation_state":"computed","paper":{"title":"Self-Supervised On-Policy Distillation for Reasoning Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Yinrong Hong, Zhiquan Tan","submitted_at":"2026-05-17T15:14:24Z","abstract_excerpt":"GRPO-style RLVR trains reasoning models from multiple on-policy attempts per prompt, but typically uses these attempts only through terminal rewards. We show that a mixed group contains a richer process signal: a correct completion is a self-generated witness of how the current policy can solve the problem, while a wrong completion provides on-policy prefixes where the policy needs correction. We introduce \\emph{Self-Supervised On-Policy Distillation} (SSOPD), which distills a teacher distribution conditioned on the shortest correct completion into prefixes of the longest wrong completion. Thi"},"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":"2605.17497","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-17T15:14:24Z","cross_cats_sorted":[],"title_canon_sha256":"2b4b1df7070041c5f7f3535dca478d946c8f1d386e6ab08bf1ac4ecdd2abbc4f","abstract_canon_sha256":"3f2de4cd88610c732db5556ff3b675f96feb066149d70b28b115ad57e565e1fa"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:04:42.251976Z","signature_b64":"wZRbcwuiyLCxPNab7E95+imojssHN8/tDZegGPEMgtyMz1Y6UqKIS7J3gGjsIpUWjeOexjbSXxOdoi0wFD2NBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f9cdd46f2ded021c928620bc5dd93f1960db15b8557032eb8cb6bad172965fbd","last_reissued_at":"2026-05-20T00:04:42.251132Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:04:42.251132Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Self-Supervised On-Policy Distillation for Reasoning Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Yinrong Hong, Zhiquan Tan","submitted_at":"2026-05-17T15:14:24Z","abstract_excerpt":"GRPO-style RLVR trains reasoning models from multiple on-policy attempts per prompt, but typically uses these attempts only through terminal rewards. We show that a mixed group contains a richer process signal: a correct completion is a self-generated witness of how the current policy can solve the problem, while a wrong completion provides on-policy prefixes where the policy needs correction. We introduce \\emph{Self-Supervised On-Policy Distillation} (SSOPD), which distills a teacher distribution conditioned on the shortest correct completion into prefixes of the longest wrong completion. Thi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.17497","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/2605.17497/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"claim_evidence","ran_at":"2026-05-19T21:41:57.673272Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T21:33:23.638878Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"bbab3aa14c8eba6e9c48f6abbe4e2a86c28a901c45f5634ee7d1593713cf38a8"},"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":"2605.17497","created_at":"2026-05-20T00:04:42.251268+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.17497v1","created_at":"2026-05-20T00:04:42.251268+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.17497","created_at":"2026-05-20T00:04:42.251268+00:00"},{"alias_kind":"pith_short_12","alias_value":"7HG5I3ZN5UBB","created_at":"2026-05-20T00:04:42.251268+00:00"},{"alias_kind":"pith_short_16","alias_value":"7HG5I3ZN5UBBZEUG","created_at":"2026-05-20T00:04:42.251268+00:00"},{"alias_kind":"pith_short_8","alias_value":"7HG5I3ZN","created_at":"2026-05-20T00:04:42.251268+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/7HG5I3ZN5UBBZEUGEC6F3WJ7DF","json":"https://pith.science/pith/7HG5I3ZN5UBBZEUGEC6F3WJ7DF.json","graph_json":"https://pith.science/api/pith-number/7HG5I3ZN5UBBZEUGEC6F3WJ7DF/graph.json","events_json":"https://pith.science/api/pith-number/7HG5I3ZN5UBBZEUGEC6F3WJ7DF/events.json","paper":"https://pith.science/paper/7HG5I3ZN"},"agent_actions":{"view_html":"https://pith.science/pith/7HG5I3ZN5UBBZEUGEC6F3WJ7DF","download_json":"https://pith.science/pith/7HG5I3ZN5UBBZEUGEC6F3WJ7DF.json","view_paper":"https://pith.science/paper/7HG5I3ZN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.17497&json=true","fetch_graph":"https://pith.science/api/pith-number/7HG5I3ZN5UBBZEUGEC6F3WJ7DF/graph.json","fetch_events":"https://pith.science/api/pith-number/7HG5I3ZN5UBBZEUGEC6F3WJ7DF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7HG5I3ZN5UBBZEUGEC6F3WJ7DF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7HG5I3ZN5UBBZEUGEC6F3WJ7DF/action/storage_attestation","attest_author":"https://pith.science/pith/7HG5I3ZN5UBBZEUGEC6F3WJ7DF/action/author_attestation","sign_citation":"https://pith.science/pith/7HG5I3ZN5UBBZEUGEC6F3WJ7DF/action/citation_signature","submit_replication":"https://pith.science/pith/7HG5I3ZN5UBBZEUGEC6F3WJ7DF/action/replication_record"}},"created_at":"2026-05-20T00:04:42.251268+00:00","updated_at":"2026-05-20T00:04:42.251268+00:00"}