{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:NGKYKIAK2YNQHDSTDVY4WVD2XY","short_pith_number":"pith:NGKYKIAK","schema_version":"1.0","canonical_sha256":"699585200ad61b038e531d71cb547abe2edccb32c3e145ebfe106541a40a44fd","source":{"kind":"arxiv","id":"2605.11458","version":2},"attestation_state":"computed","paper":{"title":"Adaptive Teacher Exposure for Self-Distillation in LLM Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Adaptive control of how much reference reasoning the teacher sees during self-distillation improves LLM performance on math tasks.","cross_cats":["cs.CL","cs.LO"],"primary_cat":"cs.AI","authors_text":"Huaibin Wang, Tiangang Zhang, Yilun Sun, Zihao Han","submitted_at":"2026-05-12T03:15:58Z","abstract_excerpt":"On-policy self-distillation has become a strong recipe for LLM reasoning, where a privileged teacher supervises the student's own rollouts while conditioning on the reference solution. A design choice shared by nearly all such methods, however, has gone unquestioned: the teacher always sees the full reference reasoning. We argue that this default itself is part of the problem and identify a teacher-side exposure mismatch: when the teacher conditions on reasoning far beyond the student's current competence, the resulting token targets become too strong to absorb. A controlled fixed-exposure swe"},"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":"2605.11458","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-12T03:15:58Z","cross_cats_sorted":["cs.CL","cs.LO"],"title_canon_sha256":"c6d1d4231c0b2deef372a95b664109c6d744977b41c04ca0d03397984278f7c7","abstract_canon_sha256":"904a1023c98c8cc7d93e5da3ce5481f1f93879c80f6c2466377bc5f15cb4ccb9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-28T01:04:42.194831Z","signature_b64":"viS9WilWCU6GnQsrvK5PHm0wC86EvWPsFMj/vlyJqEYUHqMNZbZzCrE2Yt0oQkJhig/g6ag1l0Z541bpfeFCAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"699585200ad61b038e531d71cb547abe2edccb32c3e145ebfe106541a40a44fd","last_reissued_at":"2026-05-28T01:04:42.194431Z","signature_status":"signed_v1","first_computed_at":"2026-05-28T01:04:42.194431Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Adaptive Teacher Exposure for Self-Distillation in LLM Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Adaptive control of how much reference reasoning the teacher sees during self-distillation improves LLM performance on math tasks.","cross_cats":["cs.CL","cs.LO"],"primary_cat":"cs.AI","authors_text":"Huaibin Wang, Tiangang Zhang, Yilun Sun, Zihao Han","submitted_at":"2026-05-12T03:15:58Z","abstract_excerpt":"On-policy self-distillation has become a strong recipe for LLM reasoning, where a privileged teacher supervises the student's own rollouts while conditioning on the reference solution. A design choice shared by nearly all such methods, however, has gone unquestioned: the teacher always sees the full reference reasoning. We argue that this default itself is part of the problem and identify a teacher-side exposure mismatch: when the teacher conditions on reasoning far beyond the student's current competence, the resulting token targets become too strong to absorb. A controlled fixed-exposure swe"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Experiments on AIME 24, AIME 25, and HMMT 25 across Qwen3-{1.7B, 4B, 8B} show that ATESD consistently outperforms competitive self-distillation and RL baselines, improving over OPSD by +0.95, +2.05, and +2.33 Average@12 points respectively.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That a lightweight Beta-policy controller, optimized via a discounted learning-progress reward on compact training-state statistics, will reliably produce exposure decisions that improve long-term student performance without introducing training instability or benchmark-specific overfitting.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"ATESD makes teacher exposure to reference reasoning a learnable control variable via a Beta-policy optimized on future student improvement, yielding gains of up to +2.33 points over fixed-exposure self-distillation on AIME and HMMT math benchmarks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Adaptive control of how much reference reasoning the teacher sees during self-distillation improves LLM performance on math tasks.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"247392f2a7a4b3ce7677691686c4452dd9cf055493a5c128ece74144e3dda313"},"source":{"id":"2605.11458","kind":"arxiv","version":2},"verdict":{"id":"280b6ff4-8e7c-40fd-a0ac-4edfcda05f18","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-13T01:58:33.413358Z","strongest_claim":"Experiments on AIME 24, AIME 25, and HMMT 25 across Qwen3-{1.7B, 4B, 8B} show that ATESD consistently outperforms competitive self-distillation and RL baselines, improving over OPSD by +0.95, +2.05, and +2.33 Average@12 points respectively.","one_line_summary":"ATESD makes teacher exposure to reference reasoning a learnable control variable via a Beta-policy optimized on future student improvement, yielding gains of up to +2.33 points over fixed-exposure self-distillation on AIME and HMMT math benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That a lightweight Beta-policy controller, optimized via a discounted learning-progress reward on compact training-state statistics, will reliably produce exposure decisions that improve long-term student performance without introducing training instability or benchmark-specific overfitting.","pith_extraction_headline":"Adaptive control of how much reference reasoning the teacher sees during self-distillation improves LLM performance on math tasks."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.11458/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"claim_evidence","ran_at":"2026-05-20T04:22:00.356800Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T12:35:50.141606Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-19T10:01:16.403452Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T08:27:17.389902Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"38adbe07a56e833b42a95bc4bdf342f3b25d1dbdf58aeae49e5fe179f68572f6"},"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":"2605.11458","created_at":"2026-05-28T01:04:42.194478+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.11458v2","created_at":"2026-05-28T01:04:42.194478+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.11458","created_at":"2026-05-28T01:04:42.194478+00:00"},{"alias_kind":"pith_short_12","alias_value":"NGKYKIAK2YNQ","created_at":"2026-05-28T01:04:42.194478+00:00"},{"alias_kind":"pith_short_16","alias_value":"NGKYKIAK2YNQHDST","created_at":"2026-05-28T01:04:42.194478+00:00"},{"alias_kind":"pith_short_8","alias_value":"NGKYKIAK","created_at":"2026-05-28T01:04:42.194478+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2605.18141","citing_title":"A Brief Overview: On-Policy Self-Distillation In Large Language Models","ref_index":10,"is_internal_anchor":true},{"citing_arxiv_id":"2605.18141","citing_title":"A Brief Overview: On-Policy Self-Distillation In Large Language Models","ref_index":10,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NGKYKIAK2YNQHDSTDVY4WVD2XY","json":"https://pith.science/pith/NGKYKIAK2YNQHDSTDVY4WVD2XY.json","graph_json":"https://pith.science/api/pith-number/NGKYKIAK2YNQHDSTDVY4WVD2XY/graph.json","events_json":"https://pith.science/api/pith-number/NGKYKIAK2YNQHDSTDVY4WVD2XY/events.json","paper":"https://pith.science/paper/NGKYKIAK"},"agent_actions":{"view_html":"https://pith.science/pith/NGKYKIAK2YNQHDSTDVY4WVD2XY","download_json":"https://pith.science/pith/NGKYKIAK2YNQHDSTDVY4WVD2XY.json","view_paper":"https://pith.science/paper/NGKYKIAK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.11458&json=true","fetch_graph":"https://pith.science/api/pith-number/NGKYKIAK2YNQHDSTDVY4WVD2XY/graph.json","fetch_events":"https://pith.science/api/pith-number/NGKYKIAK2YNQHDSTDVY4WVD2XY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NGKYKIAK2YNQHDSTDVY4WVD2XY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NGKYKIAK2YNQHDSTDVY4WVD2XY/action/storage_attestation","attest_author":"https://pith.science/pith/NGKYKIAK2YNQHDSTDVY4WVD2XY/action/author_attestation","sign_citation":"https://pith.science/pith/NGKYKIAK2YNQHDSTDVY4WVD2XY/action/citation_signature","submit_replication":"https://pith.science/pith/NGKYKIAK2YNQHDSTDVY4WVD2XY/action/replication_record"}},"created_at":"2026-05-28T01:04:42.194478+00:00","updated_at":"2026-05-28T01:04:42.194478+00:00"}