{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:4GBUZ74GE2NJVF7JLEETLG45SI","short_pith_number":"pith:4GBUZ74G","schema_version":"1.0","canonical_sha256":"e1834cff86269a9a97e95909359b9d9212c4c07ecc4dbe4142983df87b60a6b4","source":{"kind":"arxiv","id":"2608.01735","version":1},"attestation_state":"computed","paper":{"title":"DAPD: Dual-Anchored Policy Distillation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Chen Tang, Encheng Su, Jianyu Wu, Shixiang Tang, Yizhou Wang","submitted_at":"2026-08-03T06:00:44Z","abstract_excerpt":"On-policy (self) distillation (OPSD) is increasingly adopted for language-model post-training. It strengthens the teacher with privileged information but can induce a privilege illusion: the student learns privilege-dependent behavior it cannot reproduce from its inference-time context, yet behaves as if the training-time privileged information remained available, ultimately degrading performance. In this paper, we identify information asymmetry between the privileged teacher and the student at inference as the root cause of this failure in OPSD. To resolve this asymmetry, we propose Dual-Anch"},"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":"2608.01735","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-08-03T06:00:44Z","cross_cats_sorted":[],"title_canon_sha256":"8cf8ddb1b4638e43f9928205cf9ad6e52b7bff1bab861ca926b2687bcaaf5ee1","abstract_canon_sha256":"912ac3e88e2097870e42c4f0afc5a5aab3d4a1a74fe15c27b50ea1960e1d97d5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-04T02:06:38.741498Z","signature_b64":"vC3lTp3hw1hGZWCAVkn6xh+ARJY0kNWsVyDVueIDhXMPPt5zELQ8vL7/gHuC/IvFLWZLXYU24NxB2Loai75MDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e1834cff86269a9a97e95909359b9d9212c4c07ecc4dbe4142983df87b60a6b4","last_reissued_at":"2026-08-04T02:06:38.739956Z","signature_status":"signed_v1","first_computed_at":"2026-08-04T02:06:38.739956Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DAPD: Dual-Anchored Policy Distillation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Chen Tang, Encheng Su, Jianyu Wu, Shixiang Tang, Yizhou Wang","submitted_at":"2026-08-03T06:00:44Z","abstract_excerpt":"On-policy (self) distillation (OPSD) is increasingly adopted for language-model post-training. It strengthens the teacher with privileged information but can induce a privilege illusion: the student learns privilege-dependent behavior it cannot reproduce from its inference-time context, yet behaves as if the training-time privileged information remained available, ultimately degrading performance. In this paper, we identify information asymmetry between the privileged teacher and the student at inference as the root cause of this failure in OPSD. To resolve this asymmetry, we propose Dual-Anch"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.01735","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/2608.01735/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":"2608.01735","created_at":"2026-08-04T02:06:38.741395+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.01735v1","created_at":"2026-08-04T02:06:38.741395+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.01735","created_at":"2026-08-04T02:06:38.741395+00:00"},{"alias_kind":"pith_short_12","alias_value":"4GBUZ74GE2NJ","created_at":"2026-08-04T02:06:38.741395+00:00"},{"alias_kind":"pith_short_16","alias_value":"4GBUZ74GE2NJVF7J","created_at":"2026-08-04T02:06:38.741395+00:00"},{"alias_kind":"pith_short_8","alias_value":"4GBUZ74G","created_at":"2026-08-04T02:06:38.741395+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/4GBUZ74GE2NJVF7JLEETLG45SI","json":"https://pith.science/pith/4GBUZ74GE2NJVF7JLEETLG45SI.json","graph_json":"https://pith.science/api/pith-number/4GBUZ74GE2NJVF7JLEETLG45SI/graph.json","events_json":"https://pith.science/api/pith-number/4GBUZ74GE2NJVF7JLEETLG45SI/events.json","paper":"https://pith.science/paper/4GBUZ74G"},"agent_actions":{"view_html":"https://pith.science/pith/4GBUZ74GE2NJVF7JLEETLG45SI","download_json":"https://pith.science/pith/4GBUZ74GE2NJVF7JLEETLG45SI.json","view_paper":"https://pith.science/paper/4GBUZ74G","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.01735&json=true","fetch_graph":"https://pith.science/api/pith-number/4GBUZ74GE2NJVF7JLEETLG45SI/graph.json","fetch_events":"https://pith.science/api/pith-number/4GBUZ74GE2NJVF7JLEETLG45SI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4GBUZ74GE2NJVF7JLEETLG45SI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4GBUZ74GE2NJVF7JLEETLG45SI/action/storage_attestation","attest_author":"https://pith.science/pith/4GBUZ74GE2NJVF7JLEETLG45SI/action/author_attestation","sign_citation":"https://pith.science/pith/4GBUZ74GE2NJVF7JLEETLG45SI/action/citation_signature","submit_replication":"https://pith.science/pith/4GBUZ74GE2NJVF7JLEETLG45SI/action/replication_record"}},"created_at":"2026-08-04T02:06:38.741395+00:00","updated_at":"2026-08-04T02:06:38.741395+00:00"}