{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:ZAGE5H3VK4SUPHAZHYA3N76KVG","short_pith_number":"pith:ZAGE5H3V","schema_version":"1.0","canonical_sha256":"c80c4e9f755725479c193e01b6ffcaa9aefe0cf8600b82b9addfb075cde41820","source":{"kind":"arxiv","id":"2607.06763","version":1},"attestation_state":"computed","paper":{"title":"Trees from Marginals: Autoregressive drafting with factorized priors","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Artur Chakhvadze, Roman Knyazhitskiy, Ryan Mathieu, Yuma Oda","submitted_at":"2026-07-07T19:48:36Z","abstract_excerpt":"Speculative decoding greatly increases the interactivity of autoregressive language models by trading off computation for extra tokens generated in a single forward pass. Factorized draft models are especially efficient because they predict future-token marginals in parallel, but their independence assumption causes acceptance rates to degrade sharply as the speculative budget grows. We analyze this limitation and introduce Weaver, a lightweight autoregressive adapter that constructs proposal trees from the top-K marginals of a factorized drafter. Weaver restores conditional dependencies betwe"},"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":"2607.06763","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-07T19:48:36Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"b0d1232b52b1a7e3e56c40210b0e13c3ce4d2190805cdd54137769f6be40b675","abstract_canon_sha256":"7af1760169b48f4e26d170ca2f6c8e6f2c5be0c265e1a220d1b1de61c951afc5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-09T00:19:25.773157Z","signature_b64":"P9yFcBOK48SnArb/rNgw9gOoeXiCnuBBCpKXJZ0rJVlYNpAxP65s9eDjJtnzj5lDr2TT6rx6YuwyFgRdNZMSBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c80c4e9f755725479c193e01b6ffcaa9aefe0cf8600b82b9addfb075cde41820","last_reissued_at":"2026-07-09T00:19:25.772669Z","signature_status":"signed_v1","first_computed_at":"2026-07-09T00:19:25.772669Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Trees from Marginals: Autoregressive drafting with factorized priors","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Artur Chakhvadze, Roman Knyazhitskiy, Ryan Mathieu, Yuma Oda","submitted_at":"2026-07-07T19:48:36Z","abstract_excerpt":"Speculative decoding greatly increases the interactivity of autoregressive language models by trading off computation for extra tokens generated in a single forward pass. Factorized draft models are especially efficient because they predict future-token marginals in parallel, but their independence assumption causes acceptance rates to degrade sharply as the speculative budget grows. We analyze this limitation and introduce Weaver, a lightweight autoregressive adapter that constructs proposal trees from the top-K marginals of a factorized drafter. Weaver restores conditional dependencies betwe"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.06763","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/2607.06763/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":"2607.06763","created_at":"2026-07-09T00:19:25.772748+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.06763v1","created_at":"2026-07-09T00:19:25.772748+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.06763","created_at":"2026-07-09T00:19:25.772748+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZAGE5H3VK4SU","created_at":"2026-07-09T00:19:25.772748+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZAGE5H3VK4SUPHAZ","created_at":"2026-07-09T00:19:25.772748+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZAGE5H3V","created_at":"2026-07-09T00:19:25.772748+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/ZAGE5H3VK4SUPHAZHYA3N76KVG","json":"https://pith.science/pith/ZAGE5H3VK4SUPHAZHYA3N76KVG.json","graph_json":"https://pith.science/api/pith-number/ZAGE5H3VK4SUPHAZHYA3N76KVG/graph.json","events_json":"https://pith.science/api/pith-number/ZAGE5H3VK4SUPHAZHYA3N76KVG/events.json","paper":"https://pith.science/paper/ZAGE5H3V"},"agent_actions":{"view_html":"https://pith.science/pith/ZAGE5H3VK4SUPHAZHYA3N76KVG","download_json":"https://pith.science/pith/ZAGE5H3VK4SUPHAZHYA3N76KVG.json","view_paper":"https://pith.science/paper/ZAGE5H3V","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.06763&json=true","fetch_graph":"https://pith.science/api/pith-number/ZAGE5H3VK4SUPHAZHYA3N76KVG/graph.json","fetch_events":"https://pith.science/api/pith-number/ZAGE5H3VK4SUPHAZHYA3N76KVG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZAGE5H3VK4SUPHAZHYA3N76KVG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZAGE5H3VK4SUPHAZHYA3N76KVG/action/storage_attestation","attest_author":"https://pith.science/pith/ZAGE5H3VK4SUPHAZHYA3N76KVG/action/author_attestation","sign_citation":"https://pith.science/pith/ZAGE5H3VK4SUPHAZHYA3N76KVG/action/citation_signature","submit_replication":"https://pith.science/pith/ZAGE5H3VK4SUPHAZHYA3N76KVG/action/replication_record"}},"created_at":"2026-07-09T00:19:25.772748+00:00","updated_at":"2026-07-09T00:19:25.772748+00:00"}