{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:IXCGSNUBFC2DMUE3PYSCHACGAM","short_pith_number":"pith:IXCGSNUB","schema_version":"1.0","canonical_sha256":"45c469368128b436509b7e24238046030ff353101fb5361c77d6acb0ce9f51be","source":{"kind":"arxiv","id":"2507.17769","version":1},"attestation_state":"computed","paper":{"title":"PolyServe: Efficient Multi-SLO Serving at Scale","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.DC","authors_text":"Arvind Krishnamurthy, Baris Kasikci, Haiyang Shi, Jiaxin Shan, Kan Zhu, Le Xu, Liguang Xie","submitted_at":"2025-07-17T05:54:42Z","abstract_excerpt":"Advances in Large Language Models (LLMs) have led to a surge of LLM-powered applications. These applications have diverse token-generation latency requirements. As a result, simply classifying workloads as latency-sensitive (LS) or best-effort (BE) overlooks the nuances within the latency-sensitive category and results in suboptimal user experiences and scheduling opportunities. However, efficiently serving requests with multiple SLO requirements poses significant challenges. First, all requests within a batch generate new tokens simultaneously, which can misalign them with their distinct SLO "},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2507.17769","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2025-07-17T05:54:42Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"e522d1296f3deb0041b474aa156690800c0eb4173cc5b6b02385ebcef4081509","abstract_canon_sha256":"0dcfc5c7150dd899d80bbeaf2d8f9ac10ba074cbd7f9a0a17a8d5b425ffb316a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:42:34.768655Z","signature_b64":"3HNz8rBroSmuXzhUdIIIWQHnEVCtMLv4SQ5Fun33+UrIbdgA3RbldK6a8McAIPgOLh59Uiv6wcfQ+xIMA4M6Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"45c469368128b436509b7e24238046030ff353101fb5361c77d6acb0ce9f51be","last_reissued_at":"2026-07-05T11:42:34.768110Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:42:34.768110Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"PolyServe: Efficient Multi-SLO Serving at Scale","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.DC","authors_text":"Arvind Krishnamurthy, Baris Kasikci, Haiyang Shi, Jiaxin Shan, Kan Zhu, Le Xu, Liguang Xie","submitted_at":"2025-07-17T05:54:42Z","abstract_excerpt":"Advances in Large Language Models (LLMs) have led to a surge of LLM-powered applications. These applications have diverse token-generation latency requirements. As a result, simply classifying workloads as latency-sensitive (LS) or best-effort (BE) overlooks the nuances within the latency-sensitive category and results in suboptimal user experiences and scheduling opportunities. However, efficiently serving requests with multiple SLO requirements poses significant challenges. First, all requests within a batch generate new tokens simultaneously, which can misalign them with their distinct SLO "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.17769","kind":"arxiv","version":1},"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/2507.17769/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2507.17769","created_at":"2026-07-05T11:42:34.768178+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.17769v1","created_at":"2026-07-05T11:42:34.768178+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.17769","created_at":"2026-07-05T11:42:34.768178+00:00"},{"alias_kind":"pith_short_12","alias_value":"IXCGSNUBFC2D","created_at":"2026-07-05T11:42:34.768178+00:00"},{"alias_kind":"pith_short_16","alias_value":"IXCGSNUBFC2DMUE3","created_at":"2026-07-05T11:42:34.768178+00:00"},{"alias_kind":"pith_short_8","alias_value":"IXCGSNUB","created_at":"2026-07-05T11:42:34.768178+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.08565","citing_title":"SMetric: Rethink LLM Scheduling for Serving Agents with Balanced Session-centric Scheduling","ref_index":60,"is_internal_anchor":true},{"citing_arxiv_id":"2606.05933","citing_title":"Beyond Greedy Chunking: SLO-Aware Sliding-Window Scheduling for LLM Inference","ref_index":30,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IXCGSNUBFC2DMUE3PYSCHACGAM","json":"https://pith.science/pith/IXCGSNUBFC2DMUE3PYSCHACGAM.json","graph_json":"https://pith.science/api/pith-number/IXCGSNUBFC2DMUE3PYSCHACGAM/graph.json","events_json":"https://pith.science/api/pith-number/IXCGSNUBFC2DMUE3PYSCHACGAM/events.json","paper":"https://pith.science/paper/IXCGSNUB"},"agent_actions":{"view_html":"https://pith.science/pith/IXCGSNUBFC2DMUE3PYSCHACGAM","download_json":"https://pith.science/pith/IXCGSNUBFC2DMUE3PYSCHACGAM.json","view_paper":"https://pith.science/paper/IXCGSNUB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.17769&json=true","fetch_graph":"https://pith.science/api/pith-number/IXCGSNUBFC2DMUE3PYSCHACGAM/graph.json","fetch_events":"https://pith.science/api/pith-number/IXCGSNUBFC2DMUE3PYSCHACGAM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IXCGSNUBFC2DMUE3PYSCHACGAM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IXCGSNUBFC2DMUE3PYSCHACGAM/action/storage_attestation","attest_author":"https://pith.science/pith/IXCGSNUBFC2DMUE3PYSCHACGAM/action/author_attestation","sign_citation":"https://pith.science/pith/IXCGSNUBFC2DMUE3PYSCHACGAM/action/citation_signature","submit_replication":"https://pith.science/pith/IXCGSNUBFC2DMUE3PYSCHACGAM/action/replication_record"}},"created_at":"2026-07-05T11:42:34.768178+00:00","updated_at":"2026-07-05T11:42:34.768178+00:00"}