{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:YNOCU5S3RHFXPQ3GVQLY4XLAQ4","short_pith_number":"pith:YNOCU5S3","canonical_record":{"source":{"id":"2406.09904","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-14T10:23:45Z","cross_cats_sorted":[],"title_canon_sha256":"8366219fc6b099f5e16f91a9e7d97a2a4deb45826101a4584928d0a37d829d1d","abstract_canon_sha256":"9679ac66604589b815b086a95fc93813ce7852646fb7d16bd5568a138e291153"},"schema_version":"1.0"},"canonical_sha256":"c35c2a765b89cb77c366ac178e5d60871d596ef47a2d72d1978eb7adceedfd99","source":{"kind":"arxiv","id":"2406.09904","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.09904","created_at":"2026-07-05T08:50:30Z"},{"alias_kind":"arxiv_version","alias_value":"2406.09904v3","created_at":"2026-07-05T08:50:30Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.09904","created_at":"2026-07-05T08:50:30Z"},{"alias_kind":"pith_short_12","alias_value":"YNOCU5S3RHFX","created_at":"2026-07-05T08:50:30Z"},{"alias_kind":"pith_short_16","alias_value":"YNOCU5S3RHFXPQ3G","created_at":"2026-07-05T08:50:30Z"},{"alias_kind":"pith_short_8","alias_value":"YNOCU5S3","created_at":"2026-07-05T08:50:30Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:YNOCU5S3RHFXPQ3GVQLY4XLAQ4","target":"record","payload":{"canonical_record":{"source":{"id":"2406.09904","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-14T10:23:45Z","cross_cats_sorted":[],"title_canon_sha256":"8366219fc6b099f5e16f91a9e7d97a2a4deb45826101a4584928d0a37d829d1d","abstract_canon_sha256":"9679ac66604589b815b086a95fc93813ce7852646fb7d16bd5568a138e291153"},"schema_version":"1.0"},"canonical_sha256":"c35c2a765b89cb77c366ac178e5d60871d596ef47a2d72d1978eb7adceedfd99","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:50:30.259795Z","signature_b64":"iQnCQbHfDF6I8Nh5OHfBSl2IEJAIzaFzP4GljK+m+ETbgdCo9LbTmKZCnif572YtnOgslZMfFysCdiIa2Wd1DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c35c2a765b89cb77c366ac178e5d60871d596ef47a2d72d1978eb7adceedfd99","last_reissued_at":"2026-07-05T08:50:30.259291Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:50:30.259291Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.09904","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:50:30Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zCCPyGEUcOe2J+L+O5KBQch93ulFmnPryqu+LfBu9rhJEpFZFdjI7ZlM97O3sTMeVKHteW6JCTW766UoNAn0DQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-10T15:23:44.872292Z"},"content_sha256":"100021cddb8530b239b543a4bf5271e8208fdc66d66b1c0e1dbd6f741c1331f4","schema_version":"1.0","event_id":"sha256:100021cddb8530b239b543a4bf5271e8208fdc66d66b1c0e1dbd6f741c1331f4"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:YNOCU5S3RHFXPQ3GVQLY4XLAQ4","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"QQQ: Quality Quattuor-Bit Quantization for Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Chao Wang, Chuan Liu, Jingyang Xiang, Lei Yu, Mincong Huang, Peng Zhang, Wei Lin, Yineng Zhang, Ying Zhang, Yujie Wang","submitted_at":"2024-06-14T10:23:45Z","abstract_excerpt":"Quantization is a proven effective method for compressing large language models. Although popular techniques like W8A8 and W4A16 effectively maintain model performance, they often fail to concurrently speed up the prefill and decoding stages of inference. W4A8 is a promising strategy to accelerate both of them while usually leads to a significant performance degradation. To address these issues, we present QQQ, a Quality Quattuor-bit Quantization method with 4-bit weights and 8-bit activations. QQQ employs adaptive smoothing and Hessian-based compensation, significantly enhancing the performan"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.09904","kind":"arxiv","version":3},"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/2406.09904/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":"2026-07-05T08:50:30Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"xjxiL3D5Z+8RAci3rG5OfoBIQA1JGtV90wpJwcJiA1/X5EaxRE0zMNrqmobYAngPa/cg1A3Ws2JCGaU2GcLcDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-10T15:23:44.872697Z"},"content_sha256":"9f9d73aca10ac924e235346f44192ab08049bc7d1931b2ccf80d61dae31d1aaa","schema_version":"1.0","event_id":"sha256:9f9d73aca10ac924e235346f44192ab08049bc7d1931b2ccf80d61dae31d1aaa"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/YNOCU5S3RHFXPQ3GVQLY4XLAQ4/bundle.json","state_url":"https://pith.science/pith/YNOCU5S3RHFXPQ3GVQLY4XLAQ4/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/YNOCU5S3RHFXPQ3GVQLY4XLAQ4/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":"2026-07-10T15:23:44Z","links":{"resolver":"https://pith.science/pith/YNOCU5S3RHFXPQ3GVQLY4XLAQ4","bundle":"https://pith.science/pith/YNOCU5S3RHFXPQ3GVQLY4XLAQ4/bundle.json","state":"https://pith.science/pith/YNOCU5S3RHFXPQ3GVQLY4XLAQ4/state.json","well_known_bundle":"https://pith.science/.well-known/pith/YNOCU5S3RHFXPQ3GVQLY4XLAQ4/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:YNOCU5S3RHFXPQ3GVQLY4XLAQ4","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":"9679ac66604589b815b086a95fc93813ce7852646fb7d16bd5568a138e291153","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-14T10:23:45Z","title_canon_sha256":"8366219fc6b099f5e16f91a9e7d97a2a4deb45826101a4584928d0a37d829d1d"},"schema_version":"1.0","source":{"id":"2406.09904","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.09904","created_at":"2026-07-05T08:50:30Z"},{"alias_kind":"arxiv_version","alias_value":"2406.09904v3","created_at":"2026-07-05T08:50:30Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.09904","created_at":"2026-07-05T08:50:30Z"},{"alias_kind":"pith_short_12","alias_value":"YNOCU5S3RHFX","created_at":"2026-07-05T08:50:30Z"},{"alias_kind":"pith_short_16","alias_value":"YNOCU5S3RHFXPQ3G","created_at":"2026-07-05T08:50:30Z"},{"alias_kind":"pith_short_8","alias_value":"YNOCU5S3","created_at":"2026-07-05T08:50:30Z"}],"graph_snapshots":[{"event_id":"sha256:9f9d73aca10ac924e235346f44192ab08049bc7d1931b2ccf80d61dae31d1aaa","target":"graph","created_at":"2026-07-05T08:50:30Z","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/2406.09904/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Quantization is a proven effective method for compressing large language models. Although popular techniques like W8A8 and W4A16 effectively maintain model performance, they often fail to concurrently speed up the prefill and decoding stages of inference. W4A8 is a promising strategy to accelerate both of them while usually leads to a significant performance degradation. To address these issues, we present QQQ, a Quality Quattuor-bit Quantization method with 4-bit weights and 8-bit activations. QQQ employs adaptive smoothing and Hessian-based compensation, significantly enhancing the performan","authors_text":"Chao Wang, Chuan Liu, Jingyang Xiang, Lei Yu, Mincong Huang, Peng Zhang, Wei Lin, Yineng Zhang, Ying Zhang, Yujie Wang","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-14T10:23:45Z","title":"QQQ: Quality Quattuor-Bit Quantization for Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.09904","kind":"arxiv","version":3},"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":"sha256:100021cddb8530b239b543a4bf5271e8208fdc66d66b1c0e1dbd6f741c1331f4","target":"record","created_at":"2026-07-05T08:50:30Z","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":"9679ac66604589b815b086a95fc93813ce7852646fb7d16bd5568a138e291153","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-14T10:23:45Z","title_canon_sha256":"8366219fc6b099f5e16f91a9e7d97a2a4deb45826101a4584928d0a37d829d1d"},"schema_version":"1.0","source":{"id":"2406.09904","kind":"arxiv","version":3}},"canonical_sha256":"c35c2a765b89cb77c366ac178e5d60871d596ef47a2d72d1978eb7adceedfd99","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c35c2a765b89cb77c366ac178e5d60871d596ef47a2d72d1978eb7adceedfd99","first_computed_at":"2026-07-05T08:50:30.259291Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:50:30.259291Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"iQnCQbHfDF6I8Nh5OHfBSl2IEJAIzaFzP4GljK+m+ETbgdCo9LbTmKZCnif572YtnOgslZMfFysCdiIa2Wd1DA==","signature_status":"signed_v1","signed_at":"2026-07-05T08:50:30.259795Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.09904","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:100021cddb8530b239b543a4bf5271e8208fdc66d66b1c0e1dbd6f741c1331f4","sha256:9f9d73aca10ac924e235346f44192ab08049bc7d1931b2ccf80d61dae31d1aaa"],"state_sha256":"9c09bf456c73cf39f8e2ff15472a0006cdbd584cfd324745ea3a048c5226509f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"deJXTWQR4/X2bKeWylqHsE9RCukPgoK84nsFiTDqjLLIEVw1a18ykO8Fk0MUQgrfndh/W5FHOiDxP1RKqwa7Bg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-10T15:23:44.874826Z","bundle_sha256":"9b8f36480eb15cfba688d04ce652cc3fc26f68ab3a60e6b53c9b62e880b6ea5d"}}