{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:HFWIRGU25HV6VQ54VLJD37DQ44","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"6c42310b5341f59a4134af0d3aabecfe82e9aa6828185e6d3780f0142fe50df7","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-12T23:29:54Z","title_canon_sha256":"4f30f4f5b2caed8e0e7f93c48f7cac6f5645ef4e28b6064a4cfcfef182ae8d1c"},"schema_version":"1.0","source":{"id":"2407.09722","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.09722","created_at":"2026-07-05T10:46:55Z"},{"alias_kind":"arxiv_version","alias_value":"2407.09722v4","created_at":"2026-07-05T10:46:55Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.09722","created_at":"2026-07-05T10:46:55Z"},{"alias_kind":"pith_short_12","alias_value":"HFWIRGU25HV6","created_at":"2026-07-05T10:46:55Z"},{"alias_kind":"pith_short_16","alias_value":"HFWIRGU25HV6VQ54","created_at":"2026-07-05T10:46:55Z"},{"alias_kind":"pith_short_8","alias_value":"HFWIRGU2","created_at":"2026-07-05T10:46:55Z"}],"graph_snapshots":[{"event_id":"sha256:07950d8257ceebeb7cc400993c76e0aebd06dc41e02b7be2c14e460000f85576","target":"graph","created_at":"2026-07-05T10:46:55Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2407.09722/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs) have achieved remarkable success across diverse tasks, yet their inference processes are hindered by substantial time and energy demands due to single-token generation at each decoding step. While previous methods such as speculative decoding mitigate these inefficiencies by producing multiple tokens per step, each token is still generated by its single-token distribution, thereby enhancing speed without improving effectiveness. In contrast, our work simultaneously enhances inference speed and improves the output effectiveness. We consider multi-token joint decodin","authors_text":"Jason Cong, Neha Prakriya, Yizhou Sun, Zifan He, Ziniu Hu, Zongyue Qin","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-12T23:29:54Z","title":"Optimized Multi-Token Joint Decoding with Auxiliary Model for LLM Inference"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.09722","kind":"arxiv","version":4},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:d8df6b7ab41910bd6bb76a49885277b46ea03ad8ac3c09d1ca2a538e7d6e0498","target":"record","created_at":"2026-07-05T10:46:55Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"6c42310b5341f59a4134af0d3aabecfe82e9aa6828185e6d3780f0142fe50df7","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-12T23:29:54Z","title_canon_sha256":"4f30f4f5b2caed8e0e7f93c48f7cac6f5645ef4e28b6064a4cfcfef182ae8d1c"},"schema_version":"1.0","source":{"id":"2407.09722","kind":"arxiv","version":4}},"canonical_sha256":"396c889a9ae9ebeac3bcaad23dfc70e73551a3a0e6e56c3cf614ece36fd6f758","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"396c889a9ae9ebeac3bcaad23dfc70e73551a3a0e6e56c3cf614ece36fd6f758","first_computed_at":"2026-07-05T10:46:55.840915Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:46:55.840915Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"SYl+TyuAVISvWnKnl5P8zxf9Du6bTR+cTKSDQBg17IWHRAUYOLR3AUa/7JGrP1c3aDDEmF9BVEur8nhg7rJgDA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:46:55.841418Z","signed_message":"canonical_sha256_bytes"},"source_id":"2407.09722","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d8df6b7ab41910bd6bb76a49885277b46ea03ad8ac3c09d1ca2a538e7d6e0498","sha256:07950d8257ceebeb7cc400993c76e0aebd06dc41e02b7be2c14e460000f85576"],"state_sha256":"8d3dcd992903de03521216a19bcf9aa9b1d836324b3ed109136e3f54b5d41bc0"}