{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:X6K7YPG67X66QPFB4KPIX4GPNJ","short_pith_number":"pith:X6K7YPG6","schema_version":"1.0","canonical_sha256":"bf95fc3cdefdfde83ca1e29e8bf0cf6a49c5e0b42c62e9f8715f58b155d0c4a0","source":{"kind":"arxiv","id":"2410.01028","version":1},"attestation_state":"computed","paper":{"title":"Draft on the Fly: Adaptive Self-Speculative Decoding using Cosine Similarity","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Boxing Chen, Ivan Kobyzev, Mehdi Rezagholizadeh, Michael R. Metel, Peng Lu","submitted_at":"2024-10-01T19:35:23Z","abstract_excerpt":"We present a simple on the fly method for faster inference of large language models. Unlike other (self-)speculative decoding techniques, our method does not require fine-tuning or black-box optimization to generate a fixed draft model, relying instead on simple rules to generate varying draft models adapted to the input context. We show empirically that our light-weight algorithm is competitive with the current SOTA for self-speculative decoding, while being a truly plug-and-play method."},"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":"2410.01028","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-01T19:35:23Z","cross_cats_sorted":[],"title_canon_sha256":"96ee2c87ae7a091da17846d99fbb9e71aa94148935f3790456ced0f34c016fac","abstract_canon_sha256":"bdee7afd89a897819ba989cd959820e0643ab30d04a934c6bfe05e1b603840ce"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:14:42.118440Z","signature_b64":"G0eEQ1ZtOsKx1oBPXqTWYpahMWA8bpz03oRz4fevWwn0D6DN1genEgcJ8uiEph/oKdV9m7xagf9pl21HfZ13Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bf95fc3cdefdfde83ca1e29e8bf0cf6a49c5e0b42c62e9f8715f58b155d0c4a0","last_reissued_at":"2026-07-05T09:14:42.118014Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:14:42.118014Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Draft on the Fly: Adaptive Self-Speculative Decoding using Cosine Similarity","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Boxing Chen, Ivan Kobyzev, Mehdi Rezagholizadeh, Michael R. Metel, Peng Lu","submitted_at":"2024-10-01T19:35:23Z","abstract_excerpt":"We present a simple on the fly method for faster inference of large language models. Unlike other (self-)speculative decoding techniques, our method does not require fine-tuning or black-box optimization to generate a fixed draft model, relying instead on simple rules to generate varying draft models adapted to the input context. We show empirically that our light-weight algorithm is competitive with the current SOTA for self-speculative decoding, while being a truly plug-and-play method."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.01028","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/2410.01028/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":"2410.01028","created_at":"2026-07-05T09:14:42.118071+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.01028v1","created_at":"2026-07-05T09:14:42.118071+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.01028","created_at":"2026-07-05T09:14:42.118071+00:00"},{"alias_kind":"pith_short_12","alias_value":"X6K7YPG67X66","created_at":"2026-07-05T09:14:42.118071+00:00"},{"alias_kind":"pith_short_16","alias_value":"X6K7YPG67X66QPFB","created_at":"2026-07-05T09:14:42.118071+00:00"},{"alias_kind":"pith_short_8","alias_value":"X6K7YPG6","created_at":"2026-07-05T09:14:42.118071+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.00535","citing_title":"DREAM-S: Speculative Decoding with Searchable Drafting and Target-Aware Refinement for Multimodal Generation","ref_index":101,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/X6K7YPG67X66QPFB4KPIX4GPNJ","json":"https://pith.science/pith/X6K7YPG67X66QPFB4KPIX4GPNJ.json","graph_json":"https://pith.science/api/pith-number/X6K7YPG67X66QPFB4KPIX4GPNJ/graph.json","events_json":"https://pith.science/api/pith-number/X6K7YPG67X66QPFB4KPIX4GPNJ/events.json","paper":"https://pith.science/paper/X6K7YPG6"},"agent_actions":{"view_html":"https://pith.science/pith/X6K7YPG67X66QPFB4KPIX4GPNJ","download_json":"https://pith.science/pith/X6K7YPG67X66QPFB4KPIX4GPNJ.json","view_paper":"https://pith.science/paper/X6K7YPG6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.01028&json=true","fetch_graph":"https://pith.science/api/pith-number/X6K7YPG67X66QPFB4KPIX4GPNJ/graph.json","fetch_events":"https://pith.science/api/pith-number/X6K7YPG67X66QPFB4KPIX4GPNJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/X6K7YPG67X66QPFB4KPIX4GPNJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/X6K7YPG67X66QPFB4KPIX4GPNJ/action/storage_attestation","attest_author":"https://pith.science/pith/X6K7YPG67X66QPFB4KPIX4GPNJ/action/author_attestation","sign_citation":"https://pith.science/pith/X6K7YPG67X66QPFB4KPIX4GPNJ/action/citation_signature","submit_replication":"https://pith.science/pith/X6K7YPG67X66QPFB4KPIX4GPNJ/action/replication_record"}},"created_at":"2026-07-05T09:14:42.118071+00:00","updated_at":"2026-07-05T09:14:42.118071+00:00"}