{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:DDMY6GIIR7VE4D2EMSPCHIKBNG","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":"4d71672d303eba2a680f9191c0a3fc761e1b27b2a95b793a51227909f7f46c4d","cross_cats_sorted":["cs.DC"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-05T05:28:09Z","title_canon_sha256":"343372395adf22dd0a13c819876cea6d6c8b3d5e741a2472d6634103da8a0db1"},"schema_version":"1.0","source":{"id":"2506.04645","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.04645","created_at":"2026-07-05T11:16:18Z"},{"alias_kind":"arxiv_version","alias_value":"2506.04645v1","created_at":"2026-07-05T11:16:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.04645","created_at":"2026-07-05T11:16:18Z"},{"alias_kind":"pith_short_12","alias_value":"DDMY6GIIR7VE","created_at":"2026-07-05T11:16:18Z"},{"alias_kind":"pith_short_16","alias_value":"DDMY6GIIR7VE4D2E","created_at":"2026-07-05T11:16:18Z"},{"alias_kind":"pith_short_8","alias_value":"DDMY6GII","created_at":"2026-07-05T11:16:18Z"}],"graph_snapshots":[{"event_id":"sha256:f7edbe06dcd30f7d83ff634e8bd26bb8965b672ac6820a06a34fcead37502cbb","target":"graph","created_at":"2026-07-05T11:16:18Z","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/2506.04645/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We develop a theoretical model that addresses the economic trade-off between cost per token versus serial token generation speed when deploying LLMs for inference at scale. Our model takes into account arithmetic, memory bandwidth, network bandwidth and latency constraints; and optimizes over different parallelism setups and batch sizes to find the ones that optimize serial inference speed at a given cost per token. We use the model to compute Pareto frontiers of serial speed versus cost per token for popular language models.","authors_text":"Ege Erdil","cross_cats":["cs.DC"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-05T05:28:09Z","title":"Inference economics of language models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.04645","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":"sha256:4f9b86d39768eed13b0ef39d1d3995a40f5ed504e5a69252d0e94d55d89cdd42","target":"record","created_at":"2026-07-05T11:16:18Z","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":"4d71672d303eba2a680f9191c0a3fc761e1b27b2a95b793a51227909f7f46c4d","cross_cats_sorted":["cs.DC"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-05T05:28:09Z","title_canon_sha256":"343372395adf22dd0a13c819876cea6d6c8b3d5e741a2472d6634103da8a0db1"},"schema_version":"1.0","source":{"id":"2506.04645","kind":"arxiv","version":1}},"canonical_sha256":"18d98f19088fea4e0f44649e23a14169a600211a97a486a5071ab4184821424b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"18d98f19088fea4e0f44649e23a14169a600211a97a486a5071ab4184821424b","first_computed_at":"2026-07-05T11:16:18.434209Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:16:18.434209Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"eMG0jtdsyxCM1TSmgOMsK0yfDmCLEZ0AHnlaPrI5rdZ9w06ueRPm7Gn7sWYdNefgDeKRWMr6LsT7MkxhF78pBw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:16:18.434649Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.04645","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4f9b86d39768eed13b0ef39d1d3995a40f5ed504e5a69252d0e94d55d89cdd42","sha256:f7edbe06dcd30f7d83ff634e8bd26bb8965b672ac6820a06a34fcead37502cbb"],"state_sha256":"d55f2a7379e600f3708f6390382de1f1f6f08c30f8ae1daa15bcc3e31ebedaad"}