{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:BMZWCNLMMQVEEGZJO7BC5FOIBQ","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":"f32a69ab6cb1d7b20f86020f8750cbb534ad99acf54f119c3bfa732eb0e1be3d","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2025-06-12T21:15:58Z","title_canon_sha256":"4b6f32d95177a2cdb2c5512c072ae820ae844a5538c5e787aa864b08fe53c74f"},"schema_version":"1.0","source":{"id":"2506.11309","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.11309","created_at":"2026-07-05T11:21:02Z"},{"alias_kind":"arxiv_version","alias_value":"2506.11309v1","created_at":"2026-07-05T11:21:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.11309","created_at":"2026-07-05T11:21:02Z"},{"alias_kind":"pith_short_12","alias_value":"BMZWCNLMMQVE","created_at":"2026-07-05T11:21:02Z"},{"alias_kind":"pith_short_16","alias_value":"BMZWCNLMMQVEEGZJ","created_at":"2026-07-05T11:21:02Z"},{"alias_kind":"pith_short_8","alias_value":"BMZWCNLM","created_at":"2026-07-05T11:21:02Z"}],"graph_snapshots":[{"event_id":"sha256:1ad2c2f1110f5ed65573d1987f669faf37763557e0ef44b8cf8a64e5633a1a0b","target":"graph","created_at":"2026-07-05T11:21:02Z","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/2506.11309/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Low-latency decoding for large language models (LLMs) is crucial for applications like chatbots and code assistants, yet generating long outputs remains slow in single-query settings. Prior work on speculative decoding (which combines a small draft model with a larger target model) and tensor parallelism has each accelerated decoding. However, conventional approaches fail to apply both simultaneously due to imbalanced compute requirements (between draft and target models), KV-cache inconsistencies, and communication overheads under small-batch tensor-parallelism. This paper introduces SwiftSpe","authors_text":"Chengquan Jiang, Haibin Lin, Henry Hoffmann, Menghan Yu, Size Zheng, Xin Liu, Ziheng Jiang, Ziyi Zhang","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2025-06-12T21:15:58Z","title":"SwiftSpec: Ultra-Low Latency LLM Decoding by Scaling Asynchronous Speculative Decoding"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.11309","kind":"arxiv","version":1},"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":"sha256:88e2d558782202fd082519f20bae0e426c80486e739b3762bde669cabb1021f2","target":"record","created_at":"2026-07-05T11:21:02Z","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":"f32a69ab6cb1d7b20f86020f8750cbb534ad99acf54f119c3bfa732eb0e1be3d","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2025-06-12T21:15:58Z","title_canon_sha256":"4b6f32d95177a2cdb2c5512c072ae820ae844a5538c5e787aa864b08fe53c74f"},"schema_version":"1.0","source":{"id":"2506.11309","kind":"arxiv","version":1}},"canonical_sha256":"0b3361356c642a421b2977c22e95c80c3906459f2972f736805d87fc3bb21d13","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0b3361356c642a421b2977c22e95c80c3906459f2972f736805d87fc3bb21d13","first_computed_at":"2026-07-05T11:21:02.380090Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:21:02.380090Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"tg4OL4HNoZv57qoOIdOsGgS731S/sQz8tjLg2ds8Nc4Xc5kqCGeCG+LXks1fG6u4MHP0RqXMU3/g31lODtwQDg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:21:02.380610Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.11309","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:88e2d558782202fd082519f20bae0e426c80486e739b3762bde669cabb1021f2","sha256:1ad2c2f1110f5ed65573d1987f669faf37763557e0ef44b8cf8a64e5633a1a0b"],"state_sha256":"61388bb72b3d609b1f2d4c027e65db528163998b4ba2834b06ce78bd2890fa17"}