{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:62RH3ZGZ27NJPEVXR6GQZSPNUW","short_pith_number":"pith:62RH3ZGZ","canonical_record":{"source":{"id":"2406.11235","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-17T06:03:13Z","cross_cats_sorted":[],"title_canon_sha256":"39b2754532591a8d64031a198945bffc2f8d29e5fc2eb3ddbcebecc9eb4f9c6e","abstract_canon_sha256":"aeeedc4a875f0f33114ce07a2500122b5883cc24ce6991b922711732c3bfaf11"},"schema_version":"1.0"},"canonical_sha256":"f6a27de4d9d7da9792b78f8d0cc9eda5b47ad4a4c66466790889fd622c240cb0","source":{"kind":"arxiv","id":"2406.11235","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.11235","created_at":"2026-07-05T11:23:16Z"},{"alias_kind":"arxiv_version","alias_value":"2406.11235v4","created_at":"2026-07-05T11:23:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.11235","created_at":"2026-07-05T11:23:16Z"},{"alias_kind":"pith_short_12","alias_value":"62RH3ZGZ27NJ","created_at":"2026-07-05T11:23:16Z"},{"alias_kind":"pith_short_16","alias_value":"62RH3ZGZ27NJPEVX","created_at":"2026-07-05T11:23:16Z"},{"alias_kind":"pith_short_8","alias_value":"62RH3ZGZ","created_at":"2026-07-05T11:23:16Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:62RH3ZGZ27NJPEVXR6GQZSPNUW","target":"record","payload":{"canonical_record":{"source":{"id":"2406.11235","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-17T06:03:13Z","cross_cats_sorted":[],"title_canon_sha256":"39b2754532591a8d64031a198945bffc2f8d29e5fc2eb3ddbcebecc9eb4f9c6e","abstract_canon_sha256":"aeeedc4a875f0f33114ce07a2500122b5883cc24ce6991b922711732c3bfaf11"},"schema_version":"1.0"},"canonical_sha256":"f6a27de4d9d7da9792b78f8d0cc9eda5b47ad4a4c66466790889fd622c240cb0","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:23:16.616940Z","signature_b64":"DruFCBUUjIiFRxccG1atkFGWL48rsGyg2jIOhFmHYKyEd1aoA2gRn8Z+JrjWFKc1UN6bqVzdPMQlr2DlrVKgCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f6a27de4d9d7da9792b78f8d0cc9eda5b47ad4a4c66466790889fd622c240cb0","last_reissued_at":"2026-07-05T11:23:16.616437Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:23:16.616437Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.11235","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":"2026-07-05T11:23:16Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"sK/4gcmD5hC2fT/el8jbuYx9m0v7BBtD3kRkOySMArIe0syvXpalPfrmToh5Wty7mApFQndTXxfSuUKs8xI4Aw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T20:08:19.142807Z"},"content_sha256":"ff02a1cc9e043b910dd25877e9b194aa26516e3a0a0bc8e4a980db5ab8c4d526","schema_version":"1.0","event_id":"sha256:ff02a1cc9e043b910dd25877e9b194aa26516e3a0a0bc8e4a980db5ab8c4d526"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:62RH3ZGZ27NJPEVXR6GQZSPNUW","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"QTIP: Quantization with Trellises and Incoherence Processing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Albert Tseng, Christopher De Sa, David Hou, Qingyao Sun","submitted_at":"2024-06-17T06:03:13Z","abstract_excerpt":"Post-training quantization (PTQ) reduces the memory footprint of LLMs by quantizing weights to low-precision datatypes. Since LLM inference is usually memory-bound, PTQ methods can improve inference throughput. Recent state-of-the-art PTQ approaches use vector quantization (VQ) to quantize multiple weights at once, which improves information utilization through better shaping. However, VQ requires a codebook with size exponential in the dimension. This limits current VQ-based PTQ works to low VQ dimensions ($\\le 8$) that in turn limit quantization quality. Here, we introduce QTIP, which instea"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.11235","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/2406.11235/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":"2026-07-05T11:23:16Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"goAdG0Ueg4kpLP9D4rb7czNbL8RGYIMOgUCUP2ICW7hE10S0e0KVY3VKufP3ZLYdtJHvrpUSHHeKH7GXfQrXAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T20:08:19.143317Z"},"content_sha256":"ef4e56dac255691f1a2144fd9c3f03992969bb2b0dd2527a59a44deb3a011e55","schema_version":"1.0","event_id":"sha256:ef4e56dac255691f1a2144fd9c3f03992969bb2b0dd2527a59a44deb3a011e55"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/62RH3ZGZ27NJPEVXR6GQZSPNUW/bundle.json","state_url":"https://pith.science/pith/62RH3ZGZ27NJPEVXR6GQZSPNUW/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/62RH3ZGZ27NJPEVXR6GQZSPNUW/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":"2026-08-09T20:08:19Z","links":{"resolver":"https://pith.science/pith/62RH3ZGZ27NJPEVXR6GQZSPNUW","bundle":"https://pith.science/pith/62RH3ZGZ27NJPEVXR6GQZSPNUW/bundle.json","state":"https://pith.science/pith/62RH3ZGZ27NJPEVXR6GQZSPNUW/state.json","well_known_bundle":"https://pith.science/.well-known/pith/62RH3ZGZ27NJPEVXR6GQZSPNUW/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:62RH3ZGZ27NJPEVXR6GQZSPNUW","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":"aeeedc4a875f0f33114ce07a2500122b5883cc24ce6991b922711732c3bfaf11","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-17T06:03:13Z","title_canon_sha256":"39b2754532591a8d64031a198945bffc2f8d29e5fc2eb3ddbcebecc9eb4f9c6e"},"schema_version":"1.0","source":{"id":"2406.11235","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.11235","created_at":"2026-07-05T11:23:16Z"},{"alias_kind":"arxiv_version","alias_value":"2406.11235v4","created_at":"2026-07-05T11:23:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.11235","created_at":"2026-07-05T11:23:16Z"},{"alias_kind":"pith_short_12","alias_value":"62RH3ZGZ27NJ","created_at":"2026-07-05T11:23:16Z"},{"alias_kind":"pith_short_16","alias_value":"62RH3ZGZ27NJPEVX","created_at":"2026-07-05T11:23:16Z"},{"alias_kind":"pith_short_8","alias_value":"62RH3ZGZ","created_at":"2026-07-05T11:23:16Z"}],"graph_snapshots":[{"event_id":"sha256:ef4e56dac255691f1a2144fd9c3f03992969bb2b0dd2527a59a44deb3a011e55","target":"graph","created_at":"2026-07-05T11:23:16Z","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/2406.11235/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Post-training quantization (PTQ) reduces the memory footprint of LLMs by quantizing weights to low-precision datatypes. Since LLM inference is usually memory-bound, PTQ methods can improve inference throughput. Recent state-of-the-art PTQ approaches use vector quantization (VQ) to quantize multiple weights at once, which improves information utilization through better shaping. However, VQ requires a codebook with size exponential in the dimension. This limits current VQ-based PTQ works to low VQ dimensions ($\\le 8$) that in turn limit quantization quality. Here, we introduce QTIP, which instea","authors_text":"Albert Tseng, Christopher De Sa, David Hou, Qingyao Sun","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-17T06:03:13Z","title":"QTIP: Quantization with Trellises and Incoherence Processing"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.11235","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:ff02a1cc9e043b910dd25877e9b194aa26516e3a0a0bc8e4a980db5ab8c4d526","target":"record","created_at":"2026-07-05T11:23:16Z","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":"aeeedc4a875f0f33114ce07a2500122b5883cc24ce6991b922711732c3bfaf11","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-17T06:03:13Z","title_canon_sha256":"39b2754532591a8d64031a198945bffc2f8d29e5fc2eb3ddbcebecc9eb4f9c6e"},"schema_version":"1.0","source":{"id":"2406.11235","kind":"arxiv","version":4}},"canonical_sha256":"f6a27de4d9d7da9792b78f8d0cc9eda5b47ad4a4c66466790889fd622c240cb0","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f6a27de4d9d7da9792b78f8d0cc9eda5b47ad4a4c66466790889fd622c240cb0","first_computed_at":"2026-07-05T11:23:16.616437Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:23:16.616437Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"DruFCBUUjIiFRxccG1atkFGWL48rsGyg2jIOhFmHYKyEd1aoA2gRn8Z+JrjWFKc1UN6bqVzdPMQlr2DlrVKgCw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:23:16.616940Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.11235","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ff02a1cc9e043b910dd25877e9b194aa26516e3a0a0bc8e4a980db5ab8c4d526","sha256:ef4e56dac255691f1a2144fd9c3f03992969bb2b0dd2527a59a44deb3a011e55"],"state_sha256":"bcf0da35f2f6c99d5dcdb8280582e1dd5659a2e42840fb245b015170620e3c0a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"eJ1rrl93B3i19XAd0G89lkl60RNkdcXPobC1CRRBRbZcR6DlWWd+Y6r1LTCIWDngd2uQvbWnJGcWNcCAeIRLAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T20:08:19.148368Z","bundle_sha256":"1ab944cbb2ba413ece31cbbec6d05f102cb43985abf59574d34d3832e20e7432"}}