{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:3OCNODIJX6HIEDDBLHLCGFLL4I","short_pith_number":"pith:3OCNODIJ","schema_version":"1.0","canonical_sha256":"db84d70d09bf8e820c6159d623156be22df0dce187aeaf4b2e8f4c1920dfedcc","source":{"kind":"arxiv","id":"2206.01861","version":1},"attestation_state":"computed","paper":{"title":"ZeroQuant: Efficient and Affordable Post-Training Quantization for Large-Scale Transformers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Conglong Li, Minjia Zhang, Reza Yazdani Aminabadi, Xiaoxia Wu, Yuxiong He, Zhewei Yao","submitted_at":"2022-06-04T00:28:21Z","abstract_excerpt":"How to efficiently serve ever-larger trained natural language models in practice has become exceptionally challenging even for powerful cloud servers due to their prohibitive memory/computation requirements. In this work, we present an efficient and affordable post-training quantization approach to compress large Transformer-based models, termed as ZeroQuant. ZeroQuant is an end-to-end quantization and inference pipeline with three main components: (1) a fine-grained hardware-friendly quantization scheme for both weight and activations; (2) a novel affordable layer-by-layer knowledge distillat"},"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":"2206.01861","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-06-04T00:28:21Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"4c53dc18a570861d28d16131313f1053c893feb671f5ccf032128e2d9516b5b0","abstract_canon_sha256":"d131cbbfb37867389ea419a664a399c3ae662b4fda326cfebf3202dc30502a2b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:28:56.984347Z","signature_b64":"vOvJVeYjeA1ofYiphHaRct6nskQzYE9uH1OdXg/FhJ0bPvQfyi2ws5J7z970s9Sao1R4NNFUoGWccFuD259uAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"db84d70d09bf8e820c6159d623156be22df0dce187aeaf4b2e8f4c1920dfedcc","last_reissued_at":"2026-07-05T04:28:56.983893Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:28:56.983893Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ZeroQuant: Efficient and Affordable Post-Training Quantization for Large-Scale Transformers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Conglong Li, Minjia Zhang, Reza Yazdani Aminabadi, Xiaoxia Wu, Yuxiong He, Zhewei Yao","submitted_at":"2022-06-04T00:28:21Z","abstract_excerpt":"How to efficiently serve ever-larger trained natural language models in practice has become exceptionally challenging even for powerful cloud servers due to their prohibitive memory/computation requirements. In this work, we present an efficient and affordable post-training quantization approach to compress large Transformer-based models, termed as ZeroQuant. ZeroQuant is an end-to-end quantization and inference pipeline with three main components: (1) a fine-grained hardware-friendly quantization scheme for both weight and activations; (2) a novel affordable layer-by-layer knowledge distillat"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2206.01861","kind":"arxiv","version":1},"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/2206.01861/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":"2206.01861","created_at":"2026-07-05T04:28:56.983943+00:00"},{"alias_kind":"arxiv_version","alias_value":"2206.01861v1","created_at":"2026-07-05T04:28:56.983943+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2206.01861","created_at":"2026-07-05T04:28:56.983943+00:00"},{"alias_kind":"pith_short_12","alias_value":"3OCNODIJX6HI","created_at":"2026-07-05T04:28:56.983943+00:00"},{"alias_kind":"pith_short_16","alias_value":"3OCNODIJX6HIEDDB","created_at":"2026-07-05T04:28:56.983943+00:00"},{"alias_kind":"pith_short_8","alias_value":"3OCNODIJ","created_at":"2026-07-05T04:28:56.983943+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.09864","citing_title":"Alignment Collapse Under KV Cache Quantization: Diagnosis and Mitigation","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2605.24754","citing_title":"Motion-Compensated Weight Compression","ref_index":65,"is_internal_anchor":false},{"citing_arxiv_id":"2306.00978","citing_title":"AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2306.14048","citing_title":"H$_2$O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Models","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2601.02455","citing_title":"Diagnostic-Driven Layer-Wise Compensation for Post-Training Quantization of Encoder-Decoder ASR Models","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2208.07339","citing_title":"LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale","ref_index":171,"is_internal_anchor":false},{"citing_arxiv_id":"2305.14314","citing_title":"QLoRA: Efficient Finetuning of Quantized LLMs","ref_index":69,"is_internal_anchor":false},{"citing_arxiv_id":"2302.01318","citing_title":"Accelerating Large Language Model Decoding with Speculative Sampling","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2210.17323","citing_title":"GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2604.13440","citing_title":"A KL Lens on Quantization: Fast, Forward-Only Sensitivity for Mixed-Precision SSM-Transformer Models","ref_index":25,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3OCNODIJX6HIEDDBLHLCGFLL4I","json":"https://pith.science/pith/3OCNODIJX6HIEDDBLHLCGFLL4I.json","graph_json":"https://pith.science/api/pith-number/3OCNODIJX6HIEDDBLHLCGFLL4I/graph.json","events_json":"https://pith.science/api/pith-number/3OCNODIJX6HIEDDBLHLCGFLL4I/events.json","paper":"https://pith.science/paper/3OCNODIJ"},"agent_actions":{"view_html":"https://pith.science/pith/3OCNODIJX6HIEDDBLHLCGFLL4I","download_json":"https://pith.science/pith/3OCNODIJX6HIEDDBLHLCGFLL4I.json","view_paper":"https://pith.science/paper/3OCNODIJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2206.01861&json=true","fetch_graph":"https://pith.science/api/pith-number/3OCNODIJX6HIEDDBLHLCGFLL4I/graph.json","fetch_events":"https://pith.science/api/pith-number/3OCNODIJX6HIEDDBLHLCGFLL4I/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3OCNODIJX6HIEDDBLHLCGFLL4I/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3OCNODIJX6HIEDDBLHLCGFLL4I/action/storage_attestation","attest_author":"https://pith.science/pith/3OCNODIJX6HIEDDBLHLCGFLL4I/action/author_attestation","sign_citation":"https://pith.science/pith/3OCNODIJX6HIEDDBLHLCGFLL4I/action/citation_signature","submit_replication":"https://pith.science/pith/3OCNODIJX6HIEDDBLHLCGFLL4I/action/replication_record"}},"created_at":"2026-07-05T04:28:56.983943+00:00","updated_at":"2026-07-05T04:28:56.983943+00:00"}