{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:DLY2RBIHXKXJLGAZIH4ULWKRKB","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":"12487bddda91833f9947204b2639181c97483fdbf334a270d5a1bdf1e67e2f89","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2024-01-15T08:01:40Z","title_canon_sha256":"e058d716e43ba299a899a98bd9e792f7f3404bf60480ea3ed8991592a9460e30"},"schema_version":"1.0","source":{"id":"2402.04925","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2402.04925","created_at":"2026-07-05T07:42:34Z"},{"alias_kind":"arxiv_version","alias_value":"2402.04925v1","created_at":"2026-07-05T07:42:34Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.04925","created_at":"2026-07-05T07:42:34Z"},{"alias_kind":"pith_short_12","alias_value":"DLY2RBIHXKXJ","created_at":"2026-07-05T07:42:34Z"},{"alias_kind":"pith_short_16","alias_value":"DLY2RBIHXKXJLGAZ","created_at":"2026-07-05T07:42:34Z"},{"alias_kind":"pith_short_8","alias_value":"DLY2RBIH","created_at":"2026-07-05T07:42:34Z"}],"graph_snapshots":[{"event_id":"sha256:2cc865fb486041e7e46cfc0e646658297a7bfe8a0026ef19a4dc1cb0d121f136","target":"graph","created_at":"2026-07-05T07:42:34Z","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/2402.04925/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In this paper, we present a novel method that reduces model inference latency during distributed deployment of Large Language Models (LLMs). Our contribution is an optimized inference deployment scheme that address the current limitations of state-of-the-art quantization kernels when used in conjunction with Tensor Parallel (TP). Our method preserves data locality in GPU memory access patterns and exploits a priori knowledge of TP to reduce global communication. We demonstrate an up to 1.81x speedup over existing methods for Llama-70B and up to 1.78x speedup for IBM WatsonX's Granite-20B MLP l","authors_text":"Adnan Hoque, Chih-Chieh Yang, Mudhakar Srivatsa, Raghu Ganti","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2024-01-15T08:01:40Z","title":"TP-Aware Dequantization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.04925","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":"sha256:5ac9f469d5707297ff8cff864118efddecf006422777cf9f7f5bcd7589b53ebd","target":"record","created_at":"2026-07-05T07:42:34Z","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":"12487bddda91833f9947204b2639181c97483fdbf334a270d5a1bdf1e67e2f89","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2024-01-15T08:01:40Z","title_canon_sha256":"e058d716e43ba299a899a98bd9e792f7f3404bf60480ea3ed8991592a9460e30"},"schema_version":"1.0","source":{"id":"2402.04925","kind":"arxiv","version":1}},"canonical_sha256":"1af1a88507baae95981941f945d951504418e3d1880f7edd4f66bb20cb81691f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"1af1a88507baae95981941f945d951504418e3d1880f7edd4f66bb20cb81691f","first_computed_at":"2026-07-05T07:42:34.047344Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:42:34.047344Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"F6WZdinz03Lzq9Znxp6WiVoozUQXAPjGggUjsMPf8MOKxclBLvbpM+H3AtULU5/LzAVs5BbG4AIWKPx4WcthDg==","signature_status":"signed_v1","signed_at":"2026-07-05T07:42:34.047771Z","signed_message":"canonical_sha256_bytes"},"source_id":"2402.04925","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5ac9f469d5707297ff8cff864118efddecf006422777cf9f7f5bcd7589b53ebd","sha256:2cc865fb486041e7e46cfc0e646658297a7bfe8a0026ef19a4dc1cb0d121f136"],"state_sha256":"b134d499a332f1cc577786d8f5106a95222b10e9f7051743bfbada33735fccb7"}