{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,5]],"date-time":"2026-06-05T05:16:41Z","timestamp":1780636601437,"version":"3.54.1"},"reference-count":21,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,6,22]],"date-time":"2025-06-22T00:00:00Z","timestamp":1750550400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,6,22]],"date-time":"2025-06-22T00:00:00Z","timestamp":1750550400000},"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":[[2025,6,22]]},"DOI":"10.1109\/dac63849.2025.11133377","type":"proceedings-article","created":{"date-parts":[[2025,9,15]],"date-time":"2025-09-15T17:35:41Z","timestamp":1757957741000},"page":"1-7","source":"Crossref","is-referenced-by-count":1,"title":["A Memory-Efficient LLM Accelerator with Q-K Correlation Prediction using Cluster-Based Associative Array for Selective KV Accessing"],"prefix":"10.1109","author":[{"given":"Zikang","family":"Zhou","sequence":"first","affiliation":[{"name":"Fudan University,State Key Laboratory of Integrated Chips and Systems,Shanghai,China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kaiqi","family":"Chen","sequence":"additional","affiliation":[{"name":"Fudan University,State Key Laboratory of Integrated Chips and Systems,Shanghai,China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xuyang","family":"Duan","sequence":"additional","affiliation":[{"name":"Fudan University,State Key Laboratory of Integrated Chips and Systems,Shanghai,China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jun","family":"Han","sequence":"additional","affiliation":[{"name":"Fudan University,State Key Laboratory of Integrated Chips and Systems,Shanghai,China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Gpt-4 technical report","year":"2024"},{"key":"ref2","article-title":"Llama 2: Open foundation and fine-tuned chat models","author":"Touvron","year":"2023"},{"key":"ref3","article-title":"Opt: Open pre-trained transformer language models","author":"Zhang","year":"2022"},{"key":"ref4","article-title":"Large language models: A survey","author":"Minaee","year":"2024"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA47549.2020.00035"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA51647.2021.00018"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1145\/3466752.3480125"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA52012.2021.00060"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1145\/3503222.3507738"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA59077.2024.00077"},{"key":"ref11","article-title":"Attention is all you need","author":"Vaswani","year":"2017","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref12","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":"ref13","article-title":"A framework for few-shot language model evaluation, 12 2023","volume":"7","author":"Gao"},{"key":"ref14","first-page":"2397","article-title":"Pythia: A suite for analyzing large language models across training and scaling","volume-title":"International Conference on Machine Learning. PMLR","author":"Biderman"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v34i05.6239"},{"key":"ref16","doi-asserted-by":"crossref","DOI":"10.18653\/v1\/D18-1260","article-title":"Can a suit of armor conduct electricity? a new dataset for open book question answering","author":"Mihaylov","year":"2018"},{"key":"ref17","article-title":"Superglue: A stickier benchmark for generalpurpose language understanding systems","volume":"32","author":"Wang","year":"2019","journal-title":"Advances in neural information processing systems"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1145\/3474381"},{"key":"ref19","article-title":"Pointer sentinel mixture models","author":"Merity","year":"2016"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1145\/2228360.2228584"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/LCA.2020.2973991"}],"event":{"name":"2025 62nd ACM\/IEEE Design Automation Conference (DAC)","location":"San Francisco, CA, USA","start":{"date-parts":[[2025,6,22]]},"end":{"date-parts":[[2025,6,25]]}},"container-title":["2025 62nd ACM\/IEEE Design Automation Conference (DAC)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11132383\/11132091\/11133377.pdf?arnumber=11133377","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,9,16]],"date-time":"2025-09-16T05:47:34Z","timestamp":1758001654000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11133377\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,6,22]]},"references-count":21,"URL":"https:\/\/doi.org\/10.1109\/dac63849.2025.11133377","relation":{},"subject":[],"published":{"date-parts":[[2025,6,22]]}}}