{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:YGIS3EGVFUSL2OAUP6ZTJMQRZP","short_pith_number":"pith:YGIS3EGV","schema_version":"1.0","canonical_sha256":"c1912d90d52d24bd38147fb334b211cbd0621972f6fde89f4d394ff899928fa9","source":{"kind":"arxiv","id":"2411.18462","version":2},"attestation_state":"computed","paper":{"title":"Draft Model Knows When to Stop: Self-Verification Speculative Decoding for Long-Form Generation","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Jiahao Xu, Rui Wang, Tian Liang, Xingyu Chen, Zhaopeng Tu, Zhiwei He, Ziyin Zhang","submitted_at":"2024-11-27T15:53:17Z","abstract_excerpt":"Conventional speculative decoding (SD) methods utilize a predefined length policy for proposing drafts, which implies the premise that the target model smoothly accepts the proposed draft tokens. However, reality deviates from this assumption: the oracle draft length varies significantly, and the fixed-length policy hardly satisfies such a requirement. Moreover, such discrepancy is further exacerbated in scenarios involving complex reasoning and long-form generation, particularly under test-time scaling for reasoning-specialized models. Through both theoretical and empirical estimation, we est"},"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":"2411.18462","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-11-27T15:53:17Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"ccb7414ccc3646d7edf285dc3f7d656774ee2a7a4405671459d88f575dc9780a","abstract_canon_sha256":"431756bb1bcd688b77c4560c763822a7710da7455b5a2060d8083c7fd58e4651"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:58:22.254813Z","signature_b64":"illrJrwAOMovn6v4qUDWbpzzjtXK471bLUukwYugwnbwIi5k6KeO1dw5i99WS+pVrtaJjp4vuN3VQ6j9q0zmCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c1912d90d52d24bd38147fb334b211cbd0621972f6fde89f4d394ff899928fa9","last_reissued_at":"2026-07-05T11:58:22.254302Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:58:22.254302Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Draft Model Knows When to Stop: Self-Verification Speculative Decoding for Long-Form Generation","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Jiahao Xu, Rui Wang, Tian Liang, Xingyu Chen, Zhaopeng Tu, Zhiwei He, Ziyin Zhang","submitted_at":"2024-11-27T15:53:17Z","abstract_excerpt":"Conventional speculative decoding (SD) methods utilize a predefined length policy for proposing drafts, which implies the premise that the target model smoothly accepts the proposed draft tokens. However, reality deviates from this assumption: the oracle draft length varies significantly, and the fixed-length policy hardly satisfies such a requirement. Moreover, such discrepancy is further exacerbated in scenarios involving complex reasoning and long-form generation, particularly under test-time scaling for reasoning-specialized models. Through both theoretical and empirical estimation, we est"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.18462","kind":"arxiv","version":2},"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/2411.18462/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":"2411.18462","created_at":"2026-07-05T11:58:22.254370+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.18462v2","created_at":"2026-07-05T11:58:22.254370+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.18462","created_at":"2026-07-05T11:58:22.254370+00:00"},{"alias_kind":"pith_short_12","alias_value":"YGIS3EGVFUSL","created_at":"2026-07-05T11:58:22.254370+00:00"},{"alias_kind":"pith_short_16","alias_value":"YGIS3EGVFUSL2OAU","created_at":"2026-07-05T11:58:22.254370+00:00"},{"alias_kind":"pith_short_8","alias_value":"YGIS3EGV","created_at":"2026-07-05T11:58:22.254370+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.20503","citing_title":"FASER: Fine-Grained Phase Management for Speculative Decoding in Dynamic LLM Serving","ref_index":57,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YGIS3EGVFUSL2OAUP6ZTJMQRZP","json":"https://pith.science/pith/YGIS3EGVFUSL2OAUP6ZTJMQRZP.json","graph_json":"https://pith.science/api/pith-number/YGIS3EGVFUSL2OAUP6ZTJMQRZP/graph.json","events_json":"https://pith.science/api/pith-number/YGIS3EGVFUSL2OAUP6ZTJMQRZP/events.json","paper":"https://pith.science/paper/YGIS3EGV"},"agent_actions":{"view_html":"https://pith.science/pith/YGIS3EGVFUSL2OAUP6ZTJMQRZP","download_json":"https://pith.science/pith/YGIS3EGVFUSL2OAUP6ZTJMQRZP.json","view_paper":"https://pith.science/paper/YGIS3EGV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.18462&json=true","fetch_graph":"https://pith.science/api/pith-number/YGIS3EGVFUSL2OAUP6ZTJMQRZP/graph.json","fetch_events":"https://pith.science/api/pith-number/YGIS3EGVFUSL2OAUP6ZTJMQRZP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YGIS3EGVFUSL2OAUP6ZTJMQRZP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YGIS3EGVFUSL2OAUP6ZTJMQRZP/action/storage_attestation","attest_author":"https://pith.science/pith/YGIS3EGVFUSL2OAUP6ZTJMQRZP/action/author_attestation","sign_citation":"https://pith.science/pith/YGIS3EGVFUSL2OAUP6ZTJMQRZP/action/citation_signature","submit_replication":"https://pith.science/pith/YGIS3EGVFUSL2OAUP6ZTJMQRZP/action/replication_record"}},"created_at":"2026-07-05T11:58:22.254370+00:00","updated_at":"2026-07-05T11:58:22.254370+00:00"}