{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:E7TRJVG7WFSS4LTONEOPP6WAPG","short_pith_number":"pith:E7TRJVG7","schema_version":"1.0","canonical_sha256":"27e714d4dfb1652e2e6e691cf7fac07991c4b8fddf15a712a9263a9521a53861","source":{"kind":"arxiv","id":"2606.01476","version":1},"attestation_state":"computed","paper":{"title":"OmniOPD: Logit-Free On-Policy Distillation via Speculative Verification","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Jiayi Liu, Lizhu Zhang, Mingyi Wang, Peng Bo, Xiangjun Fan, Yifan Wu, Yuhang Zhou, Zhuokai Zhao","submitted_at":"2026-05-31T22:31:15Z","abstract_excerpt":"On-Policy Distillation (OPD) trains a student model on its own generative trajectories under dense token-level feedback from a stronger teacher, mitigating both the off-policy distribution shift of Supervised Fine-Tuning (SFT) and the sparse credit assignment of Reinforcement Learning (RL). However, standard OPD faces two coupled limitations. First, it requires direct access to the teacher's token-level logits, excluding a broad class of capable proprietary models from serving as teachers. Second, the token-level logit signal itself is brittle, depending on a narrow overlap of plausible next t"},"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":"2606.01476","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-31T22:31:15Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"0ffc4dd4fcaf1bff6690faee16eb3b34220d9faa4d73bbb20d1cd476c009c93a","abstract_canon_sha256":"5e331b9268ee6385809a2c580a07c23bc076f4067693a53da0d9fe9ef86f7115"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T02:04:34.356689Z","signature_b64":"bnC+hsBTZe3vHfxeJZdcI55XxTkwgBCvPRJGAufkIdPmDAvzZTGoE+ONPdbqKS36h3T78ZEwTEQ1YhHL7f6jAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"27e714d4dfb1652e2e6e691cf7fac07991c4b8fddf15a712a9263a9521a53861","last_reissued_at":"2026-06-02T02:04:34.356230Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T02:04:34.356230Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"OmniOPD: Logit-Free On-Policy Distillation via Speculative Verification","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Jiayi Liu, Lizhu Zhang, Mingyi Wang, Peng Bo, Xiangjun Fan, Yifan Wu, Yuhang Zhou, Zhuokai Zhao","submitted_at":"2026-05-31T22:31:15Z","abstract_excerpt":"On-Policy Distillation (OPD) trains a student model on its own generative trajectories under dense token-level feedback from a stronger teacher, mitigating both the off-policy distribution shift of Supervised Fine-Tuning (SFT) and the sparse credit assignment of Reinforcement Learning (RL). However, standard OPD faces two coupled limitations. First, it requires direct access to the teacher's token-level logits, excluding a broad class of capable proprietary models from serving as teachers. Second, the token-level logit signal itself is brittle, depending on a narrow overlap of plausible next t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.01476","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/2606.01476/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":"2606.01476","created_at":"2026-06-02T02:04:34.356296+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.01476v1","created_at":"2026-06-02T02:04:34.356296+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.01476","created_at":"2026-06-02T02:04:34.356296+00:00"},{"alias_kind":"pith_short_12","alias_value":"E7TRJVG7WFSS","created_at":"2026-06-02T02:04:34.356296+00:00"},{"alias_kind":"pith_short_16","alias_value":"E7TRJVG7WFSS4LTO","created_at":"2026-06-02T02:04:34.356296+00:00"},{"alias_kind":"pith_short_8","alias_value":"E7TRJVG7","created_at":"2026-06-02T02:04:34.356296+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/E7TRJVG7WFSS4LTONEOPP6WAPG","json":"https://pith.science/pith/E7TRJVG7WFSS4LTONEOPP6WAPG.json","graph_json":"https://pith.science/api/pith-number/E7TRJVG7WFSS4LTONEOPP6WAPG/graph.json","events_json":"https://pith.science/api/pith-number/E7TRJVG7WFSS4LTONEOPP6WAPG/events.json","paper":"https://pith.science/paper/E7TRJVG7"},"agent_actions":{"view_html":"https://pith.science/pith/E7TRJVG7WFSS4LTONEOPP6WAPG","download_json":"https://pith.science/pith/E7TRJVG7WFSS4LTONEOPP6WAPG.json","view_paper":"https://pith.science/paper/E7TRJVG7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.01476&json=true","fetch_graph":"https://pith.science/api/pith-number/E7TRJVG7WFSS4LTONEOPP6WAPG/graph.json","fetch_events":"https://pith.science/api/pith-number/E7TRJVG7WFSS4LTONEOPP6WAPG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/E7TRJVG7WFSS4LTONEOPP6WAPG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/E7TRJVG7WFSS4LTONEOPP6WAPG/action/storage_attestation","attest_author":"https://pith.science/pith/E7TRJVG7WFSS4LTONEOPP6WAPG/action/author_attestation","sign_citation":"https://pith.science/pith/E7TRJVG7WFSS4LTONEOPP6WAPG/action/citation_signature","submit_replication":"https://pith.science/pith/E7TRJVG7WFSS4LTONEOPP6WAPG/action/replication_record"}},"created_at":"2026-06-02T02:04:34.356296+00:00","updated_at":"2026-06-02T02:04:34.356296+00:00"}