paper-with-me

홈 › Papers

Q-Filters: Leveraging QK Geometry for Efficient KV Cache Compression

2025-03-04 · Nathan Godey, Alessio Devoto, Yu Zhao, Simone Scardapane, Pasquale Minervini, Éric de la Clergerie, Benoît Sagot

Autoregressive language models rely on a Key-Value (KV) Cache, which avoids re-computing past hidden states during generation, making it faster. As model sizes and context lengths grow, the KV Cache becomes a significant memory bottleneck, which calls for compression methods that limit its size during generation. In this paper, we discover surprising properties of Query (Q) and Key (K) vectors that allow us to efficiently approximate attention scores without computing the attention maps. We propose Q-Filters, a training-free KV Cache compression method that filters out less crucial Key-Value pairs based on a single context-agnostic projection. Contrarily to many alternatives, Q-Filters is compatible with FlashAttention, as it does not require direct access to attention weights. Experimental results in long-context settings demonstrate that Q-Filters is competitive with attention-based compression methods such as SnapKV in retrieval tasks while consistently outperforming efficient compression schemes such as Streaming-LLM in generation setups. Notably, Q-Filters achieves a 99% accuracy in the needle-in-a-haystack task with a x32 compression level while reducing the generation perplexity drop by up to 65% in text generation compared to Streaming-LLM.

📄 PDF Abstract BibTeX arXiv:2503.02812

Code (1)

nathangodey/qfilters 공식 구현 pytorch

Tasks

Text Generation

Methods 이 논문이 사용한 방법론

Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Attention 설명 없음

Similar Papers 제목 키워드 기반

StreamCacheVGGT: Streaming Visual Geometry Transformers with Robust Scoring and Hybrid Cache Compression

2026-04-16 · Xuanyi Liu, Chunan Yu, Deyi Ji, Qi Zhu 외 arxiv

Reconstructing dense 3D geometry from continuous video streams requires stable inference under a constant memory budget. Existing $O(1)$ frameworks primarily rely on a ``pure eviction'' paradigm, which suffers from signi…

XStreamVGGT: Extremely Memory-Efficient Streaming Vision Geometry Grounded Transformer with KV Cache Compression

2026-01-03 · Zunhai Su, Weihao Ye, Hansen Feng, Keyu Fan 외 arxiv

Learning-based 3D visual geometry models have benefited substantially from large-scale transformers. Among these, StreamVGGT leverages frame-wise causal attention for strong streaming reconstruction, but suffers from unb…

ST-Lite: Training-Free KV Cache Compression with Spatio-Trajectory Guidance for Long-Horizon GUI Agents

2026-02-27 · Bowen Zhou, Zhou Xu, Wanli Li, Jingyu Xiao 외 arxiv

Training-free KV cache compression is essential for deploying vision-language GUI agents under memory and latency constraints, yet existing methods are designed for generic language workloads and ignore the distinctive s…

CacheGen: KV Cache Compression and Streaming for Fast Large Language Model Serving

2023-10-11 · YuHan Liu, Hanchen Li, Yihua Cheng, Siddhant Ray 외

As large language models (LLMs) take on complex tasks, their inputs are supplemented with longer contexts that incorporate domain knowledge. Yet using long contexts is challenging, as nothing can be generated until the w…

Language ModelingLanguage ModellingLarge Language ModelQuantization

Leveraging Filter Correlations for Deep Model Compression

2018-11-26 · Pravendra Singh, Vinay Kumar Verma, Piyush Rai, Vinay P. Namboodiri

We present a filter correlation based model compression approach for deep convolutional neural networks. Our approach iteratively identifies pairs of filters with the largest pairwise correlations and drops one of the fi…

modelModel Compressionobject-detectionObject Detection