The paper investigates whether key‑value (KV) cache eviction strategies should vary across Transformer layers. By combining existing eviction methods in different layer configurations and profiling their performance, the authors find that heterogeneous, layer‑wise routing consistently outperforms homogeneous policies on LongBench tasks. Even with a fixed set of methods, the placement of each method strongly influences overall quality, and a single well‑chosen route surpasses all nine standalone baselines across multiple cache budgets.
By Chao Fei, Kaihua Liang, Hanzhi Hu, Hongcheng Guo, Jian Weng, Marco Canini, Panos Kalnis
arXiv:2608. 07001v1 Announce Type: new Abstract: As large language models (LLMs) process increasingly long contexts, KV cache storage and repeated access have become a major bottleneck.
By Haolin Tian, Yuzhe Liu, Tonghan Wang
arXiv:2607. 01520v1 Announce Type: new Abstract: Transformer inference on long sequences is expensive because softmax attention repeatedly reads from a large KV cache.
By Lukas Haverbeck, Carmen Amo Alonso, Andres Felipe Posada-Moreno, Sebastian Trimpe, Marco Pavone
GroupKV is a lightweight hierarchical KV cache management system designed for long‑context diffusion large language model (dLLM) inference. It partitions the context into contiguous groups and uses coarse‑to‑fine sparse selection, cross‑layer consistency for predictive prefetching, and a staleness correction mechanism to keep the cache coherent amid dynamic KV updates. The approach also incorporates streaming prefill to lower peak memory usage, achieving up to 48× longer serviceable context, 3.73× faster inference in offload‑based settings, and competitive task accuracy.
By Jinhao Wang, Zhexin Hu, Kangjie Zhou, Xin Zhou, Fangfang Liu
arXiv:2608. 08684v1 Announce Type: cross Abstract: Long-context LLM inference is bottlenecked by KV cache memory, yet distributing a limited cache budget across layers remains challenging.
By Dongjie Xu, Kai Qian, Julius, Weijie Shi, Yuxuan Sun, Minghua Tang, Fenglei Jin, Hanchi Dong, Jiajie Xu
arXiv:2604. 21335v3 Announce Type: replace Abstract: Transformer inference often requires a large KV cache, especially for long-context language modeling and multimodal generation.
By Wei Jiang, Wei Wang
HeadWiseKV is a training‑free framework that compresses the residual global key–value caches of hybrid long‑context language models by assigning each physical KV head a static, multilevel history window. It formulates cache allocation as a restricted operational rate–distortion problem and uses the SeqCalib algorithm to generate per‑head residency policies that account for interactions across layers. In evaluations on four hybrid models, HeadWiseKV preserves near‑full‑KV quality while reducing peak device memory usage by 8.59% at a 112K context length and extending the largest verified context from 114K to 161K.
By Renjie Xie, Juncheng Yang, Aoting Hu, Mingxi Zhang, Liyao Wu, Zheheng Hong, Wei Xu
arXiv:2607. 06519v1 Announce Type: new Abstract: Long-context LLM inference is increasingly limited by the memory and bandwidth cost of KV caches, yet aggressive compression can remove the layer-specific evidence needed for retrieval and multi-step reasoning.
By Anna C\'ordoba, Adam Puente Tercero, Nerea Angulo Hijo, Mar Linares Tercero, Julia Barrientos, Ainhoa Miranda, Jes\'us Olivera
arXiv:2609.37988v1 Announce Type: new
Abstract: As the context size of text processed with an LLM grows, the size of KV caches can outstrip the memory allocated for the original model weights. This i...
By Joao Monteiro, Louis B\'ethune, Anastasiia Filippova, Sonia Laguna, David Grangier, Marco Cuturi
arXiv:2502. 16886v4 Announce Type: replace-cross Abstract: To reduce memory consumption during LLM inference, a handful of methods have been proposed for KV cache pruning.
By Xuanfan Ni, Liyan Xu, Chenyang Lyu, Longyue Wang, Mo Yu, Lemao Liu, Fandong Meng, Jie Zhou, Piji Li
arXiv:2607. 22389v1 Announce Type: cross Abstract: With the rapid adoption of long-context large language models (LLMs), the continuously growing KV cache during decoding has become the critical memory bottleneck.
By Chao Fang, Jun Yin, Man Shi, Marian Verhelst
Long-context inference retains a growing key--value (KV) cache during decoding, which consumes substantial GPU memory and can reduce generation throughput. This bottleneck remains in hybrid language m...