{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:QYCTOENEBWCVUU77WPRYA65JPB","short_pith_number":"pith:QYCTOENE","schema_version":"1.0","canonical_sha256":"86053711a40d855a53ffb3e3807ba9784dff466d51a5cb07ae2976ace3b7250e","source":{"kind":"arxiv","id":"2206.09557","version":4},"attestation_state":"computed","paper":{"title":"LUT-GEMM: Quantized Matrix Multiplication based on LUTs for Efficient Inference in Large-Scale Generative Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.DC","authors_text":"Baeseong Park, Beomseok Kwon, Byeongwook Kim, Dongsoo Lee, Gunho Park, Jeonghoon Kim, Minsub Kim, Se Jung Kwon, Sungjae Lee, Youngjoo Lee","submitted_at":"2022-06-20T03:48:17Z","abstract_excerpt":"Recent advances in self-supervised learning and the Transformer architecture have significantly improved natural language processing (NLP), achieving remarkably low perplexity. However, the growing size of NLP models introduces a memory wall problem during the generation phase. To mitigate this issue, recent efforts have focused on quantizing model weights to sub-4-bit precision while preserving full precision for activations, resulting in practical speed-ups during inference on a single GPU. However, these improvements primarily stem from reduced memory movement, which necessitates a resource"},"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":"2206.09557","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2022-06-20T03:48:17Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"3bcffa3f9c57c4dcc6b26cc14b2bca7ddb3cb43d3315ee8747601bea30f9852f","abstract_canon_sha256":"93205bc0b161ffc7554581ae24b705caf329420a81b51ed08688882ca1e464d0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:02:41.224779Z","signature_b64":"3ZnmdVBXhPaxoro5zVIhzZsBQCeBw7HECcM3i43acYPw/LEoXpH1L+Lu2cRLXRAGVPo457PMIotEYoaxYyVABQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"86053711a40d855a53ffb3e3807ba9784dff466d51a5cb07ae2976ace3b7250e","last_reissued_at":"2026-07-05T08:02:41.224259Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:02:41.224259Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LUT-GEMM: Quantized Matrix Multiplication based on LUTs for Efficient Inference in Large-Scale Generative Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.DC","authors_text":"Baeseong Park, Beomseok Kwon, Byeongwook Kim, Dongsoo Lee, Gunho Park, Jeonghoon Kim, Minsub Kim, Se Jung Kwon, Sungjae Lee, Youngjoo Lee","submitted_at":"2022-06-20T03:48:17Z","abstract_excerpt":"Recent advances in self-supervised learning and the Transformer architecture have significantly improved natural language processing (NLP), achieving remarkably low perplexity. However, the growing size of NLP models introduces a memory wall problem during the generation phase. To mitigate this issue, recent efforts have focused on quantizing model weights to sub-4-bit precision while preserving full precision for activations, resulting in practical speed-ups during inference on a single GPU. However, these improvements primarily stem from reduced memory movement, which necessitates a resource"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2206.09557","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/2206.09557/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":"2206.09557","created_at":"2026-07-05T08:02:41.224334+00:00"},{"alias_kind":"arxiv_version","alias_value":"2206.09557v4","created_at":"2026-07-05T08:02:41.224334+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2206.09557","created_at":"2026-07-05T08:02:41.224334+00:00"},{"alias_kind":"pith_short_12","alias_value":"QYCTOENEBWCV","created_at":"2026-07-05T08:02:41.224334+00:00"},{"alias_kind":"pith_short_16","alias_value":"QYCTOENEBWCVUU77","created_at":"2026-07-05T08:02:41.224334+00:00"},{"alias_kind":"pith_short_8","alias_value":"QYCTOENE","created_at":"2026-07-05T08:02:41.224334+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":12,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.01607","citing_title":"MxGLUT: A Reconfigurable LUT-Centric Broadcast Dataflow Accelerator for Mixed-Precision GEMM","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2606.28438","citing_title":"When AI Reviews Its Own Code: Recursive Self-Training Collapse in Code LLMs","ref_index":112,"is_internal_anchor":false},{"citing_arxiv_id":"2306.00978","citing_title":"AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration","ref_index":25,"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":17,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15626","citing_title":"IO-SVD: Input-Output Whitened SVD for Adaptive-Rank LLM Compression","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2512.06443","citing_title":"Vec-LUT: Vector Table Lookup for Parallel Ultra-Low-Bit LLM Inference on Edge Devices","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2603.16105","citing_title":"Frequency Matters: Fast Model-Agnostic Data Curation for Pruning and Quantization","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2208.07339","citing_title":"LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale","ref_index":153,"is_internal_anchor":false},{"citing_arxiv_id":"2604.25183","citing_title":"Hardware Generation and Exploration of Lookup Table-Based Accelerators for 1.58-bit LLM Inference","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2305.14314","citing_title":"QLoRA: Efficient Finetuning of Quantized LLMs","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04058","citing_title":"MP-ISMoE: Mixed-Precision Interactive Side Mixture-of-Experts for Efficient Transfer Learning","ref_index":120,"is_internal_anchor":false},{"citing_arxiv_id":"2210.17323","citing_title":"GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers","ref_index":13,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QYCTOENEBWCVUU77WPRYA65JPB","json":"https://pith.science/pith/QYCTOENEBWCVUU77WPRYA65JPB.json","graph_json":"https://pith.science/api/pith-number/QYCTOENEBWCVUU77WPRYA65JPB/graph.json","events_json":"https://pith.science/api/pith-number/QYCTOENEBWCVUU77WPRYA65JPB/events.json","paper":"https://pith.science/paper/QYCTOENE"},"agent_actions":{"view_html":"https://pith.science/pith/QYCTOENEBWCVUU77WPRYA65JPB","download_json":"https://pith.science/pith/QYCTOENEBWCVUU77WPRYA65JPB.json","view_paper":"https://pith.science/paper/QYCTOENE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2206.09557&json=true","fetch_graph":"https://pith.science/api/pith-number/QYCTOENEBWCVUU77WPRYA65JPB/graph.json","fetch_events":"https://pith.science/api/pith-number/QYCTOENEBWCVUU77WPRYA65JPB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QYCTOENEBWCVUU77WPRYA65JPB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QYCTOENEBWCVUU77WPRYA65JPB/action/storage_attestation","attest_author":"https://pith.science/pith/QYCTOENEBWCVUU77WPRYA65JPB/action/author_attestation","sign_citation":"https://pith.science/pith/QYCTOENEBWCVUU77WPRYA65JPB/action/citation_signature","submit_replication":"https://pith.science/pith/QYCTOENEBWCVUU77WPRYA65JPB/action/replication_record"}},"created_at":"2026-07-05T08:02:41.224334+00:00","updated_at":"2026-07-05T08:02:41.224334+00:00"}