{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,7]],"date-time":"2026-07-07T15:51:21Z","timestamp":1783439481954,"version":"3.54.6"},"reference-count":39,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,3,31]],"date-time":"2025-03-31T00:00:00Z","timestamp":1743379200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,3,31]],"date-time":"2025-03-31T00:00:00Z","timestamp":1743379200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,3,31]]},"DOI":"10.23919\/date64628.2025.10992912","type":"proceedings-article","created":{"date-parts":[[2025,5,21]],"date-time":"2025-05-21T17:36:35Z","timestamp":1747848995000},"page":"1-7","source":"Crossref","is-referenced-by-count":4,"title":["Cocktail: Chunk-Adaptive Mixed-Precision Quantization for Long-Context LLM Inference"],"prefix":"10.23919","author":[{"given":"Wei","family":"Tao","sequence":"first","affiliation":[{"name":"Huazhong University of Science and Technology,Wuhan,China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bin","family":"Zhang","sequence":"additional","affiliation":[{"name":"Huazhong University of Science and Technology,Wuhan,China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiaoyang","family":"Qu","sequence":"additional","affiliation":[{"name":"Ping An Technology (Shenzhen) Co., Ltd,Shenzhen,China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jiguang","family":"Wan","sequence":"additional","affiliation":[{"name":"Huazhong University of Science and Technology,Wuhan,China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jianzong","family":"Wang","sequence":"additional","affiliation":[{"name":"Ping An Technology (Shenzhen) Co., Ltd,Shenzhen,China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.23919\/DATE58400.2024.10546707"},{"key":"ref2","first-page":"1324","article-title":"Large language models for healthcare data augmentation: An example on patient-trial matching","volume-title":"AMIA Annual Symposium Proceedings","volume":"2023","author":"Yuan","year":"2023"},{"key":"ref3","author":"Sun","year":"2023","journal-title":"Test: Text prototype aligned embedding to activate llm\u2019s ability for time series"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.23919\/DATE58400.2024.10546707"},{"key":"ref5","author":"Chen","year":"2023","journal-title":"Extending context window of large language models via positional interpolation"},{"key":"ref6","author":"Peng","year":"2023","journal-title":"Yarn: Efficient context window extension of large language models"},{"key":"ref7","author":"Shazeer","year":"2019","journal-title":"Fast transformer decoding: One write-head is all you need"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.emnlp-main.298"},{"key":"ref9","author":"Cai","year":"2024","journal-title":"Medusa: Simple llm inference acceleration framework with multiple decoding heads"},{"key":"ref10","article-title":"H2o: Heavy-hitter oracle for efficient generative inference of large language models","volume":"36","author":"Zhang","year":"2024","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref11","author":"Xiao","year":"2023","journal-title":"Efficient streaming language models with attention sinks"},{"key":"ref12","author":"Han","year":"2023","journal-title":"Lm-infinite: Simple on-the-fly length generalization for large language models"},{"key":"ref13","first-page":"16344","article-title":"Flashattention: Fast and memory-efficient exact attention with io-awareness","volume":"35","author":"Dao","year":"2022","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1145\/3600006.3613165"},{"key":"ref15","author":"Dao","year":"2023","journal-title":"Flashattention-2: Faster attention with better parallelism and work partitioning"},{"key":"ref16","author":"Liu","year":"2024","journal-title":"Kivi: A tuning-free asymmetric 2bit quantization for kv cache"},{"key":"ref17","first-page":"31094","article-title":"Flexgen: High-throughput generative inference of large language models with a single gpu","volume-title":"International Conference on Machine Learning","author":"Sheng","year":"2023"},{"key":"ref18","first-page":"196","article-title":"Atom: Low-bit quantization for efficient and accurate llm serving","volume-title":"Proceedings of Machine Learning and Systems","volume":"6","author":"Zhao","year":"2024"},{"key":"ref19","author":"Lin","year":"2024","journal-title":"Qserve: W4a8kv4 quantization and system co-design for efficient llm serving"},{"key":"ref20","author":"Kim","year":"2023","journal-title":"Squeezellm: Dense-and-sparse quantization"},{"key":"ref21","author":"Yang","year":"2024","journal-title":"No token left behind: Reliable kv cache compression via importance-aware mixed precision quantization"},{"key":"ref22","author":"Hooper","year":"2024","journal-title":"Kvquant: Towards 10 million context length llm inference with kv cache quantization"},{"key":"ref23","volume-title":"Qaq: Quality adaptive quantization for llm kv cache","author":"Dong","year":"2024"},{"key":"ref24","article-title":"Attention is all you need","author":"Vaswani","year":"2017","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref25","first-page":"38087","article-title":"Smoothquant: Accurate and efficient post-training quantization for large language models","volume-title":"International Conference on Machine Learning","author":"Xiao","year":"2023"},{"key":"ref26","author":"Frantar","year":"2022","journal-title":"Gptq: Accurate post-training quantization for generative pretrained transformers"},{"key":"ref27","first-page":"35:30318","article-title":"Gpt3. int8 (): 8-bit matrix multiplication for transformers at scale","author":"Dettmers","year":"2022","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref28","author":"Ge","year":"2023","journal-title":"Model tells you what to discard: Adaptive kv cache compression for llms"},{"key":"ref29","author":"Dettmers","year":"2023","journal-title":"Spqr: A sparse-quantized representation for near-lossless llm weight compression"},{"key":"ref30","first-page":"87","article-title":"Awq: Activation-aware weight quantization for on-device llm compression and acceleration","volume-title":"Proceedings of Machine Learning and Systems","volume":"6","author":"Lin","year":"2024"},{"key":"ref31","first-page":"9459","article-title":"Retrieval-augmented generation for knowledge-intensive nlp tasks","volume":"33","author":"Lewis","year":"2020","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref32","author":"Izacard","year":"2021","journal-title":"Unsupervised dense information retrieval with contrastive learning"},{"key":"ref33","author":"Touvron","year":"2023","journal-title":"Llama 2: Open foundation and fine-tuned chat models"},{"key":"ref34","author":"Jiang","year":"2023","journal-title":"Mistral 7b. arXiv preprint"},{"key":"ref35","article-title":"How long can context length of open-source llms truly promise","author":"Li","year":"2023","journal-title":"NeurIPS 2023 Workshop on Instruction Tuning and Instruction Following"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.acl-long.172"},{"key":"ref37","volume-title":"New and improved embedding model","year":"2022"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1145\/3404835.3463238"},{"key":"ref39","author":"Wu","year":"2022","journal-title":"Memorizing transformers"}],"event":{"name":"2025 Design, Automation & Test in Europe Conference (DATE)","location":"Lyon, France","start":{"date-parts":[[2025,3,31]]},"end":{"date-parts":[[2025,4,2]]}},"container-title":["2025 Design, Automation & Test in Europe Conference (DATE)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/10992638\/10992588\/10992912.pdf?arnumber=10992912","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,5,22]],"date-time":"2025-05-22T05:32:48Z","timestamp":1747891968000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10992912\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,3,31]]},"references-count":39,"URL":"https:\/\/doi.org\/10.23919\/date64628.2025.10992912","relation":{},"subject":[],"published":{"date-parts":[[2025,3,31]]}}}