Spend Bits Where Queries Look: KV Cache Vector Quantization with Attention-Preserving Transforms

• URL: http://arxiv.org/abs/2608.04074v1
• Date: Tue, 04 Aug 2026 16:10:59 GMT
• Title: Spend Bits Where Queries Look: KV Cache Vector Quantization with Attention-Preserving Transforms
• Abstract summary: Long-context LLM decoding reads the key-value (KV) cache at every step. Loading it takes longer than computing attention over it, so throughput is bandwidth-bound.<n>The challenge is to reduce cache size while preserving the attention products, keeping reconstruction cheap, and using a fixed per-token bit count.<n>In this paper, we formulate KV cache quantization as a transform coding problem in which distortion is the error in the attention products.
• Score: 40.94986644010638
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Long-context LLM decoding reads the key-value (KV) cache at every step. Loading it takes longer than computing attention over it, so throughput is bandwidth-bound. Hence, reducing the cache size can raise both decoding speed and serving capacity. The challenge is to reduce cache size while preserving the attention products, keeping reconstruction cheap, and using a fixed per-token bit count. At two bits per element, the most competitive methods rely on orthogonal transforms. However, existing techniques are either data-oblivious or use the query statistics without deriving the transform from a distortion criterion. Moreover, they rely on transforms built on top of random or Hadamard rotations, which equalize variances across entries rather than compacting energy, and fixed-width scalar quantizers, which are suboptimal at low rates. In this paper, we formulate KV cache quantization as a transform coding problem in which distortion is the error in the attention products. We derive closed-form optimal transforms for keys and values from calibration statistics, under a high-resolution model. We show that the optimal key transform is not orthogonal and satisfies a generalized Parseval relation: the attention-aware distortion becomes mean-squared error (MSE) in the transform domain. Thus, we can use MSE-optimal vector quantizers applied directly to the transformed key coefficients. To meet the fixed-width layout requirement, we show that grouping coefficients into equal-volume partitions makes equal-size codebooks attain the variable-rate optimum under the same high-resolution model. At two bits per element, our method, termed NOVA-KV, recovers most of the long-context retrieval accuracy lost by scalar quantization methods at comparable throughput.

Related papers

• Structured Multidimensional Representation Learning for Large Language Models [0.0]
  Transformer architectures achieve state-of-the-art performance across a wide range of pattern recognition and natural language processing tasks.<n>We introduce a structured spectral factorization of the embedding space based on the L-product for third-order tensors.<n>We prove that the proposed L-Transformer is spectrally equivalent to p parallel Transformers operating on reduced embeddings.
  arXiv  Detail & Related papers  (2026-03-05T22:34:45Z)
• WUSH: Near-Optimal Adaptive Transforms for LLM Quantization [52.77441224845925]
  Quantization to low bitwidth is a standard approach for deploying large language models.<n>A few extreme weights and activations stretch the dynamic range and reduce the effective resolution of the quantizer.<n>We derive, for the first time, closed-form optimal linear blockwise transforms for joint weight-activation quantization.
  arXiv  Detail & Related papers  (2025-11-30T16:17:34Z)
• KV-Latent: Dimensional-level KV Cache Reduction with Frequency-aware Rotary Positional Embedding [72.12756830560217]
  Large language models (LLMs) based on Transformer Decoders have become the preferred choice for conversational generative AI.<n>Despite the overall superiority of the Decoder architecture, the gradually increasing Key-Value cache during inference has emerged as a primary efficiency bottleneck.<n>By down-sampling the Key-Value vector dimensions into a latent space, we can significantly reduce the KV Cache footprint and improve inference speed.
  arXiv  Detail & Related papers  (2025-07-15T12:52:12Z)
• TurboQuant: Online Vector Quantization with Near-optimal Distortion Rate [13.14434628836727]
  Vector quantization aims to quantize high-dimensional Euclidean vectors while minimizing distortion in their geometric structure.<n>We propose TurboQuant to address both mean-squared error (MSE) and inner product distortion.<n>Our data-oblivious algorithms, suitable for online applications, achieve near-optimal distortion rates.
  arXiv  Detail & Related papers  (2025-04-28T15:05:35Z)
• Quantize What Counts: More for Keys, Less for Values [63.51476878610841]
  Large Language Models (LLMs) suffer inference-time memory bottlenecks dominated by the attention Key-Value ( KV) cache.<n>This paper proposes two theorems that anchor mixed-precision KV quantization in the intrinsic geometry of Transformer models.
  arXiv  Detail & Related papers  (2025-02-20T22:24:27Z)
• Variable-size Symmetry-based Graph Fourier Transforms for image compression [65.7352685872625]
  We propose a new family of Symmetry-based Graph Fourier Transforms of variable sizes into a coding framework. Our proposed algorithm generates symmetric graphs on the grid by adding specific symmetrical connections between nodes. Experiments show that SBGFTs outperform the primary transforms integrated in the explicit Multiple Transform Selection.
  arXiv  Detail & Related papers  (2024-11-24T13:00:44Z)
• Residual vector quantization for KV cache compression in large language model [2.3094645821058735]
  KV cache compression methods have mainly relied on scalar quantization techniques to reduce the memory requirements during decoding. In this work, we apply residual vector quantization, which has been widely used for high fidelity audio compression, to compress KV cache in large language models (LLM) We learn the codebook using exponential moving average and there are no other learnable parameters including the input and output projections normally used in a vector quantization set up.
  arXiv  Detail & Related papers  (2024-10-21T07:20:41Z)
• Unlocking Data-free Low-bit Quantization with Matrix Decomposition for KV Cache Compression [87.5604418100301]
  Key-value( KV) caching is an important technique to accelerate the inference of large language models. Existing methods often compromise precision or require extra data for calibration. We introduce textbfDecoQuant, a novel data-free low-bit quantization technique based on tensor decomposition methods.
  arXiv  Detail & Related papers  (2024-05-21T08:35:10Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.