{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:4OIO474N7Y47XRTIUGEN7BRNSK","short_pith_number":"pith:4OIO474N","schema_version":"1.0","canonical_sha256":"e390ee7f8dfe39fbc668a188df862d92a993feecec851c6a99d6ec753f75202e","source":{"kind":"arxiv","id":"2409.17066","version":2},"attestation_state":"computed","paper":{"title":"VPTQ: Extreme Low-bit Vector Post-Training Quantization for Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Cheng Li, Jicheng Wen, Li Lyna Zhang, Mao Yang, Shengyu Ye, Ting Cao, Yang Wang, Yifei Liu","submitted_at":"2024-09-25T16:25:45Z","abstract_excerpt":"Scaling model size significantly challenges the deployment and inference of Large Language Models (LLMs). Due to the redundancy in LLM weights, recent research has focused on pushing weight-only quantization to extremely low-bit (even down to 2 bits). It reduces memory requirements, optimizes storage costs, and decreases memory bandwidth needs during inference. However, due to numerical representation limitations, traditional scalar-based weight quantization struggles to achieve such extreme low-bit. Recent research on Vector Quantization (VQ) for LLMs has demonstrated the potential for extrem"},"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":"2409.17066","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-09-25T16:25:45Z","cross_cats_sorted":[],"title_canon_sha256":"1f70b4c4ebedd6a871cfc2e81e83adb6a98f595a31c84bb43e6aab412af6e6ce","abstract_canon_sha256":"5523cc411e3593e850f9a573e033e11c29ac269c5df54ac7154b2e5929bfb7ac"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:23:59.576712Z","signature_b64":"9pSHyhDZw1jLr1i2HHlconpBa04lBfutEKK9ehNlrdAWIgqszL9hlC3UkQpT2M91LFpbs2VHIBdMqBKJtVhOBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e390ee7f8dfe39fbc668a188df862d92a993feecec851c6a99d6ec753f75202e","last_reissued_at":"2026-07-05T09:23:59.576161Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:23:59.576161Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"VPTQ: Extreme Low-bit Vector Post-Training Quantization for Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Cheng Li, Jicheng Wen, Li Lyna Zhang, Mao Yang, Shengyu Ye, Ting Cao, Yang Wang, Yifei Liu","submitted_at":"2024-09-25T16:25:45Z","abstract_excerpt":"Scaling model size significantly challenges the deployment and inference of Large Language Models (LLMs). Due to the redundancy in LLM weights, recent research has focused on pushing weight-only quantization to extremely low-bit (even down to 2 bits). It reduces memory requirements, optimizes storage costs, and decreases memory bandwidth needs during inference. However, due to numerical representation limitations, traditional scalar-based weight quantization struggles to achieve such extreme low-bit. Recent research on Vector Quantization (VQ) for LLMs has demonstrated the potential for extrem"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.17066","kind":"arxiv","version":2},"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/2409.17066/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":"2409.17066","created_at":"2026-07-05T09:23:59.576221+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.17066v2","created_at":"2026-07-05T09:23:59.576221+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.17066","created_at":"2026-07-05T09:23:59.576221+00:00"},{"alias_kind":"pith_short_12","alias_value":"4OIO474N7Y47","created_at":"2026-07-05T09:23:59.576221+00:00"},{"alias_kind":"pith_short_16","alias_value":"4OIO474N7Y47XRTI","created_at":"2026-07-05T09:23:59.576221+00:00"},{"alias_kind":"pith_short_8","alias_value":"4OIO474N","created_at":"2026-07-05T09:23:59.576221+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":9,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.04050","citing_title":"LiftQuant: Continuous Bit-Width LLM via Dimensional Lifting and Projection","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2605.24019","citing_title":"MGVQ: Synergizing Multi-dimensional Sensitivity-Aware and Gradient-Hessian Fusion for Vector Quantization","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.24144","citing_title":"EVA: Accelerating LLM Decoding via an Efficient Vector Quantization Architecture","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2606.04050","citing_title":"LiftQuant: Continuous Bit-Width LLM via Dimensional Lifting and Projection","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2411.05961","citing_title":"Aligned Vector Quantization for Edge-Cloud Collabrative Vision-Language Models","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2602.04163","citing_title":"BPDQ: Bit-Plane Decomposition Quantization on a Variable Grid for Large Language Models","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15551","citing_title":"Characterizing Learning in Deep Neural Networks using Tractable Algorithmic Complexity Analysis","ref_index":66,"is_internal_anchor":false},{"citing_arxiv_id":"2506.12040","citing_title":"BTC-LLM: Efficient Sub-1-Bit LLM Quantization via Learnable Transformation and Binary Codebook","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19167","citing_title":"LBLLM: Lightweight Binarization of Large Language Models via Three-Stage Distillation","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4OIO474N7Y47XRTIUGEN7BRNSK","json":"https://pith.science/pith/4OIO474N7Y47XRTIUGEN7BRNSK.json","graph_json":"https://pith.science/api/pith-number/4OIO474N7Y47XRTIUGEN7BRNSK/graph.json","events_json":"https://pith.science/api/pith-number/4OIO474N7Y47XRTIUGEN7BRNSK/events.json","paper":"https://pith.science/paper/4OIO474N"},"agent_actions":{"view_html":"https://pith.science/pith/4OIO474N7Y47XRTIUGEN7BRNSK","download_json":"https://pith.science/pith/4OIO474N7Y47XRTIUGEN7BRNSK.json","view_paper":"https://pith.science/paper/4OIO474N","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.17066&json=true","fetch_graph":"https://pith.science/api/pith-number/4OIO474N7Y47XRTIUGEN7BRNSK/graph.json","fetch_events":"https://pith.science/api/pith-number/4OIO474N7Y47XRTIUGEN7BRNSK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4OIO474N7Y47XRTIUGEN7BRNSK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4OIO474N7Y47XRTIUGEN7BRNSK/action/storage_attestation","attest_author":"https://pith.science/pith/4OIO474N7Y47XRTIUGEN7BRNSK/action/author_attestation","sign_citation":"https://pith.science/pith/4OIO474N7Y47XRTIUGEN7BRNSK/action/citation_signature","submit_replication":"https://pith.science/pith/4OIO474N7Y47XRTIUGEN7BRNSK/action/replication_record"}},"created_at":"2026-07-05T09:23:59.576221+00:00","updated_at":"2026-07-05T09:23:59.576221+00:00"}