{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/think-thinner-key-cache-by-query-driven","title":"ThinK: Thinner Key Cache by Query-Driven Pruning","arxiv_id":"2407.21018","date":"2024-07-30","proceeding":null,"authors":["Yuhui Xu","Zhanming Jie","Hanze Dong","Lei Wang","Xudong Lu","Aojun Zhou","Amrita Saha","Caiming Xiong","Doyen Sahoo"],"abstract":"Large Language Models (LLMs) have revolutionized the field of natural language processing, achieving unprecedented performance across a variety of applications. However, their increased computational and memory demands present significant challenges, especially when handling long sequences. This paper focuses on the long-context scenario, addressing the inefficiencies in KV cache memory consumption during inference. Unlike existing approaches that optimize the memory based on the sequence length, we identify substantial redundancy in the channel dimension of the KV cache, as indicated by an uneven magnitude distribution and a low-rank structure in the attention weights. In response, we propose ThinK, a novel query-dependent KV cache pruning method designed to minimize attention weight loss while selectively pruning the least significant channels. Our approach not only maintains or enhances model accuracy but also achieves a reduction in KV cache memory costs by over 20% compared with vanilla KV cache eviction and quantization methods. For instance, ThinK integrated with KIVI can achieve a 2.8x reduction in peak memory usage while maintaining nearly the same quality, enabling up to a 5x increase in batch size when using a single GPU. Extensive evaluations on the LLaMA and Mistral models across various long-sequence datasets verified the efficiency of ThinK, establishing a new baseline algorithm for efficient LLM deployment without compromising performance.","url_abs":"https://arxiv.org/abs/2407.21018v2","url_pdf":"https://arxiv.org/pdf/2407.21018v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[],"tasks":[{"task_slug":null,"task_name":"GPU"},{"task_slug":"quantization","task_name":"Quantization"}],"methods":[{"method_slug":"attention","method_name":"Attention"},{"method_slug":"llama","method_name":"LLaMA"},{"method_slug":"pruning","method_name":"Pruning"},{"method_slug":"softmax","method_name":"Softmax"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2407.21018","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2407.21018"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"deterministic:regex_extraction","url":"https://github.com/SalesforceAIResearch/ThinK","reach":null}],"summary":{"ran_fixture":1},"by_repo_kind":{"found_in_text":{"samples":1,"ran":1,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"b3ef4115883f4afe","entry":"key_pruner_query_driven","repo":"SalesforceAIResearch/ThinK","repo_kind":"found_in_text","path":"ThinK_kivi/models/llama_kivi_think.py","file_url":"https://github.com/SalesforceAIResearch/ThinK/blob/HEAD/ThinK_kivi/models/llama_kivi_think.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"b3ef4115883f4afe"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}