# ZigZagkv: Dynamic KV Cache Compression for Long-context Modeling based on Layer Uncertainty

arXiv: https://arxiv.org/abs/2412.09036 · HTML: https://mlsanity.com/paper/2412.09036/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Placed in KV cache optimization, one of 12 directions in Efficient LLM inference and serving.
- Builds on 17 of its 26 references that are in the map.
- First posted on arXiv 12 Dec 2024.

## Placed in

- Efficient LLM inference and serving / [KV cache optimization](https://mlsanity.com/inference/kv-cache-optimization/)

## Builds on (most cited first, 17 of 17)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Efficient Streaming Language Models with Attention Sinks](https://mlsanity.com/paper/2309.17453/) (2023)
- [Code Llama: Open Foundation Models for Code](https://mlsanity.com/paper/2308.12950/) (2023)
- [LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding](https://mlsanity.com/paper/2308.14508/) (2023)
- [Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Steps](https://mlsanity.com/paper/2011.01060/) (2020)
- [SnapKV: LLM Knows What You are Looking for Before Generation](https://mlsanity.com/paper/2404.14469/) (2024)
- [How Can We Know When Language Models Know? On the Calibration of Language Models for Question Answering](https://mlsanity.com/paper/2012.00955/) (2020)
- [Scissorhands: Exploiting the Persistence of Importance Hypothesis for LLM KV Cache Compression at Test Time](https://mlsanity.com/paper/2305.17118/) (2023)
- [PyramidKV: Dynamic KV Cache Compression based on Pyramidal Information Funneling](https://mlsanity.com/paper/2406.02069/) (2024)
- [Data Engineering for Scaling Language Models to 128K Context](https://mlsanity.com/paper/2402.10171/) (2024)
- [PyramidInfer: Pyramid KV Cache Compression for High-throughput LLM Inference](https://mlsanity.com/paper/2405.12532/) (2024)
- [Towards the Law of Capacity Gap in Distilling Language Models](https://mlsanity.com/paper/2311.07052/) (2023)
- [On the Efficacy of Eviction Policy for Key-Value Constrained Generative Language Model Inference](https://mlsanity.com/paper/2402.06262/) (2024)
- [Understanding the RoPE Extensions of Long-Context LLMs: An Attention Perspective](https://mlsanity.com/paper/2406.13282/) (2024)
- [MoDification: Mixture of Depths Made Easy](https://mlsanity.com/paper/2410.14268/) (2024)

## Built on by (most cited first, 7 of 7)

- [KV Cache Compression for Inference Efficiency in LLMs: A Review](https://mlsanity.com/paper/2508.06297/) (2025)
- [Make Each Token Count: Towards Improving Long-Context Performance with KV Cache Eviction](https://mlsanity.com/paper/2605.09649/) (2026)
- [SmallKV: Small Model Assisted Compensation of KV Cache Compression for Efficient LLM Inference](https://mlsanity.com/paper/2508.02751/) (2025)
- [SemShareKV: Efficient KVCache Sharing for Semantically Similar Prompts via Token-Level LSH Matching](https://mlsanity.com/paper/2509.24832/) (2025)
- [State Rank Dynamics in Linear Attention LLMs](https://mlsanity.com/paper/2602.02195/) (2026)
- [CompilerKV: Risk-Adaptive KV Compression via Offline Experience Compilation](https://mlsanity.com/paper/2602.08686/) (2026)
- [CONF-KV: Confidence-Aware KV Cache Eviction with Mixed-Precision Storage for Long-Horizon LLM](https://mlsanity.com/paper/2605.24786/) (2026)

## Abstract (arXiv)

Large Language models (LLMs) have become a research hotspot. To accelerate the inference of LLMs, storing computed caches in memory has become the standard technique. However, as the inference length increases, growing KV caches might lead to out-of-memory issues. Many existing methods address this issue through KV cache compression, primarily by preserving key tokens throughout all layers to reduce information loss. Most of them allocate a uniform budget size for each layer to retain. However, we observe that the minimum budget sizes needed to retain essential information vary across layers and models based on the perspectives of attention and hidden state output. Building on this observation, this paper proposes a simple yet effective KV cache compression method that leverages layer uncertainty to allocate budget size for each layer. Experimental results show that the proposed method can reduce memory usage of the KV caches to only ~20% when compared to Full KV inference while achieving nearly lossless performance.