{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:IUWLQCSGFUWDECUSQPTIYEQ6ZQ","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":"64ccf35c7aba444f8cbfad5c68d0d28474036eceec2be88b3b13303bd2170317","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-06T20:18:34Z","title_canon_sha256":"08a017afb2a6c36ac126bb9975114dc2b9d20b4cd0f7795d48dcf78ffa08a6ca"},"schema_version":"1.0","source":{"id":"2511.04791","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2511.04791","created_at":"2026-06-02T01:03:37Z"},{"alias_kind":"arxiv_version","alias_value":"2511.04791v2","created_at":"2026-06-02T01:03:37Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2511.04791","created_at":"2026-06-02T01:03:37Z"},{"alias_kind":"pith_short_12","alias_value":"IUWLQCSGFUWD","created_at":"2026-06-02T01:03:37Z"},{"alias_kind":"pith_short_16","alias_value":"IUWLQCSGFUWDECUS","created_at":"2026-06-02T01:03:37Z"},{"alias_kind":"pith_short_8","alias_value":"IUWLQCSG","created_at":"2026-06-02T01:03:37Z"}],"graph_snapshots":[{"event_id":"sha256:b75fa96c0204ce3c904422e72c58c01cb24fbec6e5e85e429b4887c57088e04a","target":"graph","created_at":"2026-06-02T01:03:37Z","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/2511.04791/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Modern LLM serving systems must sustain high throughput while meeting strict latency SLOs across two distinct inference phases: compute-intensive prefill and memory-bound decode phases. Existing approaches either (1) aggregate both phases on shared GPUs, leading to interference between prefill and decode phases, which degrades Time-Between-Tokens (TBT); or (2) disaggregate the two phases across GPUs, improving latency but wasting resources through duplicated models and KV cache transfers. We present DuetServe, a unified LLM serving framework that achieves disaggregation-level isolation within ","authors_text":"Chaoyi Jiang, Daniel Wong, Hossein Entezari Zarch, Lei Gao, Mark Hill, Murali Annavaram","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-06T20:18:34Z","title":"DuetServe: Harmonizing Prefill and Decode for LLM Serving via Adaptive GPU Multiplexing"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2511.04791","kind":"arxiv","version":2},"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:d4f119b0e6eea485938c83833e4c050da59771e927a15ba6ce1f765943aa8710","target":"record","created_at":"2026-06-02T01:03:37Z","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":"64ccf35c7aba444f8cbfad5c68d0d28474036eceec2be88b3b13303bd2170317","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-06T20:18:34Z","title_canon_sha256":"08a017afb2a6c36ac126bb9975114dc2b9d20b4cd0f7795d48dcf78ffa08a6ca"},"schema_version":"1.0","source":{"id":"2511.04791","kind":"arxiv","version":2}},"canonical_sha256":"452cb80a462d2c320a9283e68c121ecc1c175502e5107e0b812b9dc7ddd880de","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"452cb80a462d2c320a9283e68c121ecc1c175502e5107e0b812b9dc7ddd880de","first_computed_at":"2026-06-02T01:03:37.951027Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-02T01:03:37.951027Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"RDRkzNrIAcsw1KW7fHShBhh+3e1IE9W0u4R/tUisyDogXYOYp2PZkjbMGqdQ1wsRjm3lVvWrwmqruSj0ngRfBA==","signature_status":"signed_v1","signed_at":"2026-06-02T01:03:37.951702Z","signed_message":"canonical_sha256_bytes"},"source_id":"2511.04791","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d4f119b0e6eea485938c83833e4c050da59771e927a15ba6ce1f765943aa8710","sha256:b75fa96c0204ce3c904422e72c58c01cb24fbec6e5e85e429b4887c57088e04a"],"state_sha256":"53f9595ac86a141ce65df653c43447bfa43aa65e001dc69e6c0d16cc408743cf"}