{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:4KFQZ4ID33B5YVETS4V5VJWJDG","short_pith_number":"pith:4KFQZ4ID","schema_version":"1.0","canonical_sha256":"e28b0cf103dec3dc5493972bdaa6c919aa6aea9e9dd4bad93208d24f3a249fe7","source":{"kind":"arxiv","id":"2605.22263","version":1},"attestation_state":"computed","paper":{"title":"Tailoring Teaching to Aptitude: Direction-Adaptive Self-Distillation for LLM Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Chaozheng Wang, Hongbin Zhang, Jinpeng Wang, Kehai Chen, Min Zhang, Yang Xiang, Youcheng Pan","submitted_at":"2026-05-21T10:07:46Z","abstract_excerpt":"On-policy self-distillation (OPSD) is an emerging LLM post-training paradigm in which the model serves as its own teacher: conditioned on privileged information such as a reference trace or hint, the same policy provides dense token-level supervision on its own rollouts. However, recent studies show that OPSD degrades complex reasoning by suppressing predictive uncertainty, which supports exploration and hypothesis revision. Our token-level analysis shows that this failure arises from applying a uniform direction of teacher supervision across tokens with different uncertainty levels: conformit"},"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":"2605.22263","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-21T10:07:46Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"08d1640b9aede080c9360fa2374dbbdb58813fb03256bbdc8cc2b28e66a221e4","abstract_canon_sha256":"8e68bfd1cc43d7be412cb9346ec62a72ddecb4eef73040c182cdecbced0fab3a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-22T01:04:35.124996Z","signature_b64":"QzWzP5DwQPoF6XnW4afyLzMi+vCpS2HSViUNMWYYn8lFEN5lkJKjZADgqVXCP8phD+/ZSdYMiW7mSkrCJzV8AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e28b0cf103dec3dc5493972bdaa6c919aa6aea9e9dd4bad93208d24f3a249fe7","last_reissued_at":"2026-05-22T01:04:35.124222Z","signature_status":"signed_v1","first_computed_at":"2026-05-22T01:04:35.124222Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Tailoring Teaching to Aptitude: Direction-Adaptive Self-Distillation for LLM Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Chaozheng Wang, Hongbin Zhang, Jinpeng Wang, Kehai Chen, Min Zhang, Yang Xiang, Youcheng Pan","submitted_at":"2026-05-21T10:07:46Z","abstract_excerpt":"On-policy self-distillation (OPSD) is an emerging LLM post-training paradigm in which the model serves as its own teacher: conditioned on privileged information such as a reference trace or hint, the same policy provides dense token-level supervision on its own rollouts. However, recent studies show that OPSD degrades complex reasoning by suppressing predictive uncertainty, which supports exploration and hypothesis revision. Our token-level analysis shows that this failure arises from applying a uniform direction of teacher supervision across tokens with different uncertainty levels: conformit"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.22263","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/2605.22263/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":"2605.22263","created_at":"2026-05-22T01:04:35.124366+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.22263v1","created_at":"2026-05-22T01:04:35.124366+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.22263","created_at":"2026-05-22T01:04:35.124366+00:00"},{"alias_kind":"pith_short_12","alias_value":"4KFQZ4ID33B5","created_at":"2026-05-22T01:04:35.124366+00:00"},{"alias_kind":"pith_short_16","alias_value":"4KFQZ4ID33B5YVET","created_at":"2026-05-22T01:04:35.124366+00:00"},{"alias_kind":"pith_short_8","alias_value":"4KFQZ4ID","created_at":"2026-05-22T01:04:35.124366+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/4KFQZ4ID33B5YVETS4V5VJWJDG","json":"https://pith.science/pith/4KFQZ4ID33B5YVETS4V5VJWJDG.json","graph_json":"https://pith.science/api/pith-number/4KFQZ4ID33B5YVETS4V5VJWJDG/graph.json","events_json":"https://pith.science/api/pith-number/4KFQZ4ID33B5YVETS4V5VJWJDG/events.json","paper":"https://pith.science/paper/4KFQZ4ID"},"agent_actions":{"view_html":"https://pith.science/pith/4KFQZ4ID33B5YVETS4V5VJWJDG","download_json":"https://pith.science/pith/4KFQZ4ID33B5YVETS4V5VJWJDG.json","view_paper":"https://pith.science/paper/4KFQZ4ID","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.22263&json=true","fetch_graph":"https://pith.science/api/pith-number/4KFQZ4ID33B5YVETS4V5VJWJDG/graph.json","fetch_events":"https://pith.science/api/pith-number/4KFQZ4ID33B5YVETS4V5VJWJDG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4KFQZ4ID33B5YVETS4V5VJWJDG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4KFQZ4ID33B5YVETS4V5VJWJDG/action/storage_attestation","attest_author":"https://pith.science/pith/4KFQZ4ID33B5YVETS4V5VJWJDG/action/author_attestation","sign_citation":"https://pith.science/pith/4KFQZ4ID33B5YVETS4V5VJWJDG/action/citation_signature","submit_replication":"https://pith.science/pith/4KFQZ4ID33B5YVETS4V5VJWJDG/action/replication_record"}},"created_at":"2026-05-22T01:04:35.124366+00:00","updated_at":"2026-05-22T01:04:35.124366+00:00"}