{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:3Z4OLEVJ6TB4273MNTN55RAFTT","short_pith_number":"pith:3Z4OLEVJ","canonical_record":{"source":{"id":"2411.02355","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-04T18:21:59Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7bd39450a6afb31213afb6b0c7309c713220e53a7124fd9bc80637df5896851a","abstract_canon_sha256":"1941cb9a60744b3650d8b0ee431fd4bd53d466100acd5905bad772a74319a536"},"schema_version":"1.0"},"canonical_sha256":"de78e592a9f4c3cd7f6c6cdbdec4059cf418972be16ab02f91b5ee58106d08eb","source":{"kind":"arxiv","id":"2411.02355","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2411.02355","created_at":"2026-05-27T01:05:32Z"},{"alias_kind":"arxiv_version","alias_value":"2411.02355v4","created_at":"2026-05-27T01:05:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.02355","created_at":"2026-05-27T01:05:32Z"},{"alias_kind":"pith_short_12","alias_value":"3Z4OLEVJ6TB4","created_at":"2026-05-27T01:05:32Z"},{"alias_kind":"pith_short_16","alias_value":"3Z4OLEVJ6TB4273M","created_at":"2026-05-27T01:05:32Z"},{"alias_kind":"pith_short_8","alias_value":"3Z4OLEVJ","created_at":"2026-05-27T01:05:32Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:3Z4OLEVJ6TB4273MNTN55RAFTT","target":"record","payload":{"canonical_record":{"source":{"id":"2411.02355","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-04T18:21:59Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7bd39450a6afb31213afb6b0c7309c713220e53a7124fd9bc80637df5896851a","abstract_canon_sha256":"1941cb9a60744b3650d8b0ee431fd4bd53d466100acd5905bad772a74319a536"},"schema_version":"1.0"},"canonical_sha256":"de78e592a9f4c3cd7f6c6cdbdec4059cf418972be16ab02f91b5ee58106d08eb","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-27T01:05:32.777600Z","signature_b64":"Du1KFGIM0AXUD7lzLMYgqKYgEbKr1Adwc0hBdd82r9rkWZwzgTRw1jlaxgu44Fp3LcRKS7LckfNgsGq81S7GDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"de78e592a9f4c3cd7f6c6cdbdec4059cf418972be16ab02f91b5ee58106d08eb","last_reissued_at":"2026-05-27T01:05:32.777051Z","signature_status":"signed_v1","first_computed_at":"2026-05-27T01:05:32.777051Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2411.02355","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-27T01:05:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"b9LPB41eS39K50l/FRsKCcxM4zu7MpIy0VtGGYtHJJadgYA8szas0ldAT0J7IY0p5CFICloX9nXD9Fg3/Zr5CQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T21:39:07.106489Z"},"content_sha256":"30e885a88be86f0b16391fed8149790997e58bba97c77014395fdd206e492e22","schema_version":"1.0","event_id":"sha256:30e885a88be86f0b16391fed8149790997e58bba97c77014395fdd206e492e22"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:3Z4OLEVJ6TB4273MNTN55RAFTT","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"\"Give Me BF16 or Give Me Death\"? Accuracy-Performance Trade-Offs in LLM Quantization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Alexandre Marques, Dan Alistarh, Eldar Kurtic, Mark Kurtz, Shubhra Pandit","submitted_at":"2024-11-04T18:21:59Z","abstract_excerpt":"Quantization is a powerful tool for accelerating large language model (LLM) inference, but the accuracy-performance trade-offs across different formats remain unclear. In this paper, we conduct the most comprehensive empirical study to date, evaluating FP8, INT8, and INT4 quantization across academic benchmarks and real-world tasks on the entire Llama-3.1 model family. Through over 500,000 evaluations, our investigation yields several key findings: (1) FP8 (W8A8-FP) is effectively lossless across all model scales, (2) well-tuned INT8 (W8A8-INT) achieves surprisingly low (1-3\\%) accuracy degrad"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.02355","kind":"arxiv","version":4},"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/2411.02355/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":"2026-05-27T01:05:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"MdGcsIMWnkpNHpb9PsRl18/GyXVEcnOaqYTNGWTB2ILxYERpWGA2I3Qg0RubQi7M8gpiQFrYU4BujWXXDCwSCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T21:39:07.107411Z"},"content_sha256":"8438f879d98893fa802b8d7dc6f3a8110c2bb31143e2ff82cf68a6e099386948","schema_version":"1.0","event_id":"sha256:8438f879d98893fa802b8d7dc6f3a8110c2bb31143e2ff82cf68a6e099386948"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3Z4OLEVJ6TB4273MNTN55RAFTT/bundle.json","state_url":"https://pith.science/pith/3Z4OLEVJ6TB4273MNTN55RAFTT/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3Z4OLEVJ6TB4273MNTN55RAFTT/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":"2026-05-30T21:39:07Z","links":{"resolver":"https://pith.science/pith/3Z4OLEVJ6TB4273MNTN55RAFTT","bundle":"https://pith.science/pith/3Z4OLEVJ6TB4273MNTN55RAFTT/bundle.json","state":"https://pith.science/pith/3Z4OLEVJ6TB4273MNTN55RAFTT/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3Z4OLEVJ6TB4273MNTN55RAFTT/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:3Z4OLEVJ6TB4273MNTN55RAFTT","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":"1941cb9a60744b3650d8b0ee431fd4bd53d466100acd5905bad772a74319a536","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-04T18:21:59Z","title_canon_sha256":"7bd39450a6afb31213afb6b0c7309c713220e53a7124fd9bc80637df5896851a"},"schema_version":"1.0","source":{"id":"2411.02355","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2411.02355","created_at":"2026-05-27T01:05:32Z"},{"alias_kind":"arxiv_version","alias_value":"2411.02355v4","created_at":"2026-05-27T01:05:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.02355","created_at":"2026-05-27T01:05:32Z"},{"alias_kind":"pith_short_12","alias_value":"3Z4OLEVJ6TB4","created_at":"2026-05-27T01:05:32Z"},{"alias_kind":"pith_short_16","alias_value":"3Z4OLEVJ6TB4273M","created_at":"2026-05-27T01:05:32Z"},{"alias_kind":"pith_short_8","alias_value":"3Z4OLEVJ","created_at":"2026-05-27T01:05:32Z"}],"graph_snapshots":[{"event_id":"sha256:8438f879d98893fa802b8d7dc6f3a8110c2bb31143e2ff82cf68a6e099386948","target":"graph","created_at":"2026-05-27T01:05:32Z","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/2411.02355/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Quantization is a powerful tool for accelerating large language model (LLM) inference, but the accuracy-performance trade-offs across different formats remain unclear. In this paper, we conduct the most comprehensive empirical study to date, evaluating FP8, INT8, and INT4 quantization across academic benchmarks and real-world tasks on the entire Llama-3.1 model family. Through over 500,000 evaluations, our investigation yields several key findings: (1) FP8 (W8A8-FP) is effectively lossless across all model scales, (2) well-tuned INT8 (W8A8-INT) achieves surprisingly low (1-3\\%) accuracy degrad","authors_text":"Alexandre Marques, Dan Alistarh, Eldar Kurtic, Mark Kurtz, Shubhra Pandit","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-04T18:21:59Z","title":"\"Give Me BF16 or Give Me Death\"? Accuracy-Performance Trade-Offs in LLM Quantization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.02355","kind":"arxiv","version":4},"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":"sha256:30e885a88be86f0b16391fed8149790997e58bba97c77014395fdd206e492e22","target":"record","created_at":"2026-05-27T01:05:32Z","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":"1941cb9a60744b3650d8b0ee431fd4bd53d466100acd5905bad772a74319a536","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-04T18:21:59Z","title_canon_sha256":"7bd39450a6afb31213afb6b0c7309c713220e53a7124fd9bc80637df5896851a"},"schema_version":"1.0","source":{"id":"2411.02355","kind":"arxiv","version":4}},"canonical_sha256":"de78e592a9f4c3cd7f6c6cdbdec4059cf418972be16ab02f91b5ee58106d08eb","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"de78e592a9f4c3cd7f6c6cdbdec4059cf418972be16ab02f91b5ee58106d08eb","first_computed_at":"2026-05-27T01:05:32.777051Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-27T01:05:32.777051Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Du1KFGIM0AXUD7lzLMYgqKYgEbKr1Adwc0hBdd82r9rkWZwzgTRw1jlaxgu44Fp3LcRKS7LckfNgsGq81S7GDA==","signature_status":"signed_v1","signed_at":"2026-05-27T01:05:32.777600Z","signed_message":"canonical_sha256_bytes"},"source_id":"2411.02355","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:30e885a88be86f0b16391fed8149790997e58bba97c77014395fdd206e492e22","sha256:8438f879d98893fa802b8d7dc6f3a8110c2bb31143e2ff82cf68a6e099386948"],"state_sha256":"911692410b075467536587d76b4dceaf2f2e51f7dbc4992de023ba633dc0860e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zQA9LdqaQAlqJJ+ncgLh5vWjTcFkEm/JQYjesmd6pr7+c4/BJv3pdQZDYeIXfNXb0KiFxze2MQAkJB2gdOWIBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-30T21:39:07.111969Z","bundle_sha256":"0eb4ad813e65f39d966fae0a52dc4d5af6344b071ce7e69903651bf983f1eb16"}}