{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:GHGZHR4SOWGX4TA3PXLCKRLDTD","short_pith_number":"pith:GHGZHR4S","canonical_record":{"source":{"id":"2403.02310","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-03-04T18:47:08Z","cross_cats_sorted":["cs.DC"],"title_canon_sha256":"fb6101eaa47cfb38dc7c3b00b9657dc6764af414e743aea0ff6f5b2ba154f349","abstract_canon_sha256":"61171ab1ac6d3cc06344843109ff0dde116bddafdf3beb2a4873e09ab5e48647"},"schema_version":"1.0"},"canonical_sha256":"31cd93c792758d7e4c1b7dd625456398e6251473abcd60e225de119b96fb4e8b","source":{"kind":"arxiv","id":"2403.02310","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.02310","created_at":"2026-07-05T08:33:01Z"},{"alias_kind":"arxiv_version","alias_value":"2403.02310v3","created_at":"2026-07-05T08:33:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.02310","created_at":"2026-07-05T08:33:01Z"},{"alias_kind":"pith_short_12","alias_value":"GHGZHR4SOWGX","created_at":"2026-07-05T08:33:01Z"},{"alias_kind":"pith_short_16","alias_value":"GHGZHR4SOWGX4TA3","created_at":"2026-07-05T08:33:01Z"},{"alias_kind":"pith_short_8","alias_value":"GHGZHR4S","created_at":"2026-07-05T08:33:01Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:GHGZHR4SOWGX4TA3PXLCKRLDTD","target":"record","payload":{"canonical_record":{"source":{"id":"2403.02310","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-03-04T18:47:08Z","cross_cats_sorted":["cs.DC"],"title_canon_sha256":"fb6101eaa47cfb38dc7c3b00b9657dc6764af414e743aea0ff6f5b2ba154f349","abstract_canon_sha256":"61171ab1ac6d3cc06344843109ff0dde116bddafdf3beb2a4873e09ab5e48647"},"schema_version":"1.0"},"canonical_sha256":"31cd93c792758d7e4c1b7dd625456398e6251473abcd60e225de119b96fb4e8b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:33:01.323735Z","signature_b64":"4oHdDkR4vyvEJKq3TolsEDvMMuoAkDpNokXngBuHSOHcgHKXmhyG25uV+9ujVvE0Gx4Pv2zvnxvrA9HGNVpKAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"31cd93c792758d7e4c1b7dd625456398e6251473abcd60e225de119b96fb4e8b","last_reissued_at":"2026-07-05T08:33:01.323232Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:33:01.323232Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2403.02310","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:33:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"cchsCnfYtkP/yMFlCrtl8qSdigOrIv9ZBG+ZBm9XJ4fBLW8DWz3qrgKQdNDrSBfxTBkthABFePcZPR1jCrw2Dg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-22T21:56:10.964272Z"},"content_sha256":"db23a15ce34e06f67e89a5290e0de204afeae92c2dc5c8d96b8173060f4ff6f4","schema_version":"1.0","event_id":"sha256:db23a15ce34e06f67e89a5290e0de204afeae92c2dc5c8d96b8173060f4ff6f4"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:GHGZHR4SOWGX4TA3PXLCKRLDTD","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Serve","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC"],"primary_cat":"cs.LG","authors_text":"Alexey Tumanov, Amey Agrawal, Ashish Panwar, Bhargav S. Gulavani, Jayashree Mohan, Nipun Kwatra, Nitin Kedia, Ramachandran Ramjee","submitted_at":"2024-03-04T18:47:08Z","abstract_excerpt":"Each LLM serving request goes through two phases. The first is prefill which processes the entire input prompt and produces the first output token and the second is decode which generates the rest of output tokens, one-at-a-time. Prefill iterations have high latency but saturate GPU compute due to parallel processing of the input prompt. In contrast, decode iterations have low latency but also low compute utilization because a decode iteration processes only a single token per request. This makes batching highly effective for decodes and consequently for overall throughput. However, batching m"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.02310","kind":"arxiv","version":3},"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/2403.02310/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:33:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"98+7n4rlP9BR2JzEXJ0kbIEdeWvJIRUh9Amn+UyURKBVm8OYsXUuCtNdIOPBYuebX8v2PDzgEj7+FodyDY4FDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-22T21:56:10.964797Z"},"content_sha256":"f277031bd228292b084d3649e455fd42c96305310b0e216c39d60cb3fb823093","schema_version":"1.0","event_id":"sha256:f277031bd228292b084d3649e455fd42c96305310b0e216c39d60cb3fb823093"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/GHGZHR4SOWGX4TA3PXLCKRLDTD/bundle.json","state_url":"https://pith.science/pith/GHGZHR4SOWGX4TA3PXLCKRLDTD/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/GHGZHR4SOWGX4TA3PXLCKRLDTD/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-22T21:56:10Z","links":{"resolver":"https://pith.science/pith/GHGZHR4SOWGX4TA3PXLCKRLDTD","bundle":"https://pith.science/pith/GHGZHR4SOWGX4TA3PXLCKRLDTD/bundle.json","state":"https://pith.science/pith/GHGZHR4SOWGX4TA3PXLCKRLDTD/state.json","well_known_bundle":"https://pith.science/.well-known/pith/GHGZHR4SOWGX4TA3PXLCKRLDTD/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:GHGZHR4SOWGX4TA3PXLCKRLDTD","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":"61171ab1ac6d3cc06344843109ff0dde116bddafdf3beb2a4873e09ab5e48647","cross_cats_sorted":["cs.DC"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-03-04T18:47:08Z","title_canon_sha256":"fb6101eaa47cfb38dc7c3b00b9657dc6764af414e743aea0ff6f5b2ba154f349"},"schema_version":"1.0","source":{"id":"2403.02310","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.02310","created_at":"2026-07-05T08:33:01Z"},{"alias_kind":"arxiv_version","alias_value":"2403.02310v3","created_at":"2026-07-05T08:33:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.02310","created_at":"2026-07-05T08:33:01Z"},{"alias_kind":"pith_short_12","alias_value":"GHGZHR4SOWGX","created_at":"2026-07-05T08:33:01Z"},{"alias_kind":"pith_short_16","alias_value":"GHGZHR4SOWGX4TA3","created_at":"2026-07-05T08:33:01Z"},{"alias_kind":"pith_short_8","alias_value":"GHGZHR4S","created_at":"2026-07-05T08:33:01Z"}],"graph_snapshots":[{"event_id":"sha256:f277031bd228292b084d3649e455fd42c96305310b0e216c39d60cb3fb823093","target":"graph","created_at":"2026-07-05T08:33:01Z","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/2403.02310/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Each LLM serving request goes through two phases. The first is prefill which processes the entire input prompt and produces the first output token and the second is decode which generates the rest of output tokens, one-at-a-time. Prefill iterations have high latency but saturate GPU compute due to parallel processing of the input prompt. In contrast, decode iterations have low latency but also low compute utilization because a decode iteration processes only a single token per request. This makes batching highly effective for decodes and consequently for overall throughput. However, batching m","authors_text":"Alexey Tumanov, Amey Agrawal, Ashish Panwar, Bhargav S. Gulavani, Jayashree Mohan, Nipun Kwatra, Nitin Kedia, Ramachandran Ramjee","cross_cats":["cs.DC"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-03-04T18:47:08Z","title":"Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Serve"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.02310","kind":"arxiv","version":3},"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:db23a15ce34e06f67e89a5290e0de204afeae92c2dc5c8d96b8173060f4ff6f4","target":"record","created_at":"2026-07-05T08:33:01Z","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":"61171ab1ac6d3cc06344843109ff0dde116bddafdf3beb2a4873e09ab5e48647","cross_cats_sorted":["cs.DC"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-03-04T18:47:08Z","title_canon_sha256":"fb6101eaa47cfb38dc7c3b00b9657dc6764af414e743aea0ff6f5b2ba154f349"},"schema_version":"1.0","source":{"id":"2403.02310","kind":"arxiv","version":3}},"canonical_sha256":"31cd93c792758d7e4c1b7dd625456398e6251473abcd60e225de119b96fb4e8b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"31cd93c792758d7e4c1b7dd625456398e6251473abcd60e225de119b96fb4e8b","first_computed_at":"2026-07-05T08:33:01.323232Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:33:01.323232Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"4oHdDkR4vyvEJKq3TolsEDvMMuoAkDpNokXngBuHSOHcgHKXmhyG25uV+9ujVvE0Gx4Pv2zvnxvrA9HGNVpKAQ==","signature_status":"signed_v1","signed_at":"2026-07-05T08:33:01.323735Z","signed_message":"canonical_sha256_bytes"},"source_id":"2403.02310","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:db23a15ce34e06f67e89a5290e0de204afeae92c2dc5c8d96b8173060f4ff6f4","sha256:f277031bd228292b084d3649e455fd42c96305310b0e216c39d60cb3fb823093"],"state_sha256":"a723a2b414a8c9b5cd1c8b7d37f05079f2845eed88ed0a0454abd3e69979a675"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fEiIdyRv6m9WD3nIP8V9ezg2i21WMxMnIhfSgSW085T9x52gUXkK+OmsRKuQgDuE7au69dpmmQ1Enz5F1+C3CQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-22T21:56:10.968566Z","bundle_sha256":"b855d0baa07c197fe76a2f083ae43541b9c42d5cb102b3f1147b7d40f88e8b84"}}