{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:MWRLQGOOWMQJMVKI5IUM5T4CZD","short_pith_number":"pith:MWRLQGOO","schema_version":"1.0","canonical_sha256":"65a2b819ceb320965548ea28cecf82c8c8bad78756a8527e87aea44c303b2817","source":{"kind":"arxiv","id":"2311.00502","version":2},"attestation_state":"computed","paper":{"title":"Efficient LLM Inference on CPUs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Bo Dong, Haihao Shen, Hanwen Chang, Hengyu Meng, Yu Luo","submitted_at":"2023-11-01T13:08:50Z","abstract_excerpt":"Large language models (LLMs) have demonstrated remarkable performance and tremendous potential across a wide range of tasks. However, deploying these models has been challenging due to the astronomical amount of model parameters, which requires a demand for large memory capacity and high memory bandwidth. In this paper, we propose an effective approach that can make the deployment of LLMs more efficiently. We support an automatic INT4 weight-only quantization flow and design a special LLM runtime with highly-optimized kernels to accelerate the LLM inference on CPUs. We demonstrate the general "},"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":"2311.00502","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-11-01T13:08:50Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"b28efea4faeea74655e52a4b16474058e341b3feec8d415d084716c7dc20b66e","abstract_canon_sha256":"59b5890c44ac7a19b685ed034e775c1011cba22620342d0756f738fa2925ace0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:21:28.718087Z","signature_b64":"TVIYdUWJ4I2G03nTu2ysK6eyaG3HD6447TvtdTRVH0pwJFoZqpaFEEMuKZ6LlQpm2/EJKVJLlud6SGsCHZYHBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"65a2b819ceb320965548ea28cecf82c8c8bad78756a8527e87aea44c303b2817","last_reissued_at":"2026-07-05T07:21:28.717585Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:21:28.717585Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Efficient LLM Inference on CPUs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Bo Dong, Haihao Shen, Hanwen Chang, Hengyu Meng, Yu Luo","submitted_at":"2023-11-01T13:08:50Z","abstract_excerpt":"Large language models (LLMs) have demonstrated remarkable performance and tremendous potential across a wide range of tasks. However, deploying these models has been challenging due to the astronomical amount of model parameters, which requires a demand for large memory capacity and high memory bandwidth. In this paper, we propose an effective approach that can make the deployment of LLMs more efficiently. We support an automatic INT4 weight-only quantization flow and design a special LLM runtime with highly-optimized kernels to accelerate the LLM inference on CPUs. We demonstrate the general "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.00502","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/2311.00502/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":"2311.00502","created_at":"2026-07-05T07:21:28.717649+00:00"},{"alias_kind":"arxiv_version","alias_value":"2311.00502v2","created_at":"2026-07-05T07:21:28.717649+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.00502","created_at":"2026-07-05T07:21:28.717649+00:00"},{"alias_kind":"pith_short_12","alias_value":"MWRLQGOOWMQJ","created_at":"2026-07-05T07:21:28.717649+00:00"},{"alias_kind":"pith_short_16","alias_value":"MWRLQGOOWMQJMVKI","created_at":"2026-07-05T07:21:28.717649+00:00"},{"alias_kind":"pith_short_8","alias_value":"MWRLQGOO","created_at":"2026-07-05T07:21:28.717649+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2501.05465","citing_title":"Small Language Models (SLMs) Can Still Pack a Punch: A survey (updated 2026)","ref_index":115,"is_internal_anchor":false},{"citing_arxiv_id":"2507.18454","citing_title":"Sandwich: Joint Configuration Search and Hot-Switching for Efficient CPU LLM Serving","ref_index":53,"is_internal_anchor":false},{"citing_arxiv_id":"2512.09427","citing_title":"ODMA: On-Demand Memory Allocation Strategy for LLM Serving on LPDDR-Class Accelerators","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11948","citing_title":"Active Imitation Learning for Thermal- and Kernel-Aware LFM Inference on 3D S-NUCA Many-Cores","ref_index":36,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MWRLQGOOWMQJMVKI5IUM5T4CZD","json":"https://pith.science/pith/MWRLQGOOWMQJMVKI5IUM5T4CZD.json","graph_json":"https://pith.science/api/pith-number/MWRLQGOOWMQJMVKI5IUM5T4CZD/graph.json","events_json":"https://pith.science/api/pith-number/MWRLQGOOWMQJMVKI5IUM5T4CZD/events.json","paper":"https://pith.science/paper/MWRLQGOO"},"agent_actions":{"view_html":"https://pith.science/pith/MWRLQGOOWMQJMVKI5IUM5T4CZD","download_json":"https://pith.science/pith/MWRLQGOOWMQJMVKI5IUM5T4CZD.json","view_paper":"https://pith.science/paper/MWRLQGOO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2311.00502&json=true","fetch_graph":"https://pith.science/api/pith-number/MWRLQGOOWMQJMVKI5IUM5T4CZD/graph.json","fetch_events":"https://pith.science/api/pith-number/MWRLQGOOWMQJMVKI5IUM5T4CZD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MWRLQGOOWMQJMVKI5IUM5T4CZD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MWRLQGOOWMQJMVKI5IUM5T4CZD/action/storage_attestation","attest_author":"https://pith.science/pith/MWRLQGOOWMQJMVKI5IUM5T4CZD/action/author_attestation","sign_citation":"https://pith.science/pith/MWRLQGOOWMQJMVKI5IUM5T4CZD/action/citation_signature","submit_replication":"https://pith.science/pith/MWRLQGOOWMQJMVKI5IUM5T4CZD/action/replication_record"}},"created_at":"2026-07-05T07:21:28.717649+00:00","updated_at":"2026-07-05T07:21:28.717649+00:00"}