# The Pitfalls of KV Cache Compression

arXiv: https://arxiv.org/abs/2510.00231 · HTML: https://mlsanity.com/paper/2510.00231/

## Key facts

- Cited by 15 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 26 of its 38 references that are in the map.
- First posted on arXiv 30 Sep 2025.

## Placed in

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

## Builds on (most cited first, 26 of 26)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Efficient Streaming Language Models with Attention Sinks](https://mlsanity.com/paper/2309.17453/) (2023)
- [LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding](https://mlsanity.com/paper/2308.14508/) (2023)
- [Instruction-Following Evaluation for Large Language Models](https://mlsanity.com/paper/2311.07911/) (2023)
- [SnapKV: LLM Knows What You are Looking for Before Generation](https://mlsanity.com/paper/2404.14469/) (2024)
- [H2O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Models](https://mlsanity.com/paper/2306.14048/) (2023)
- [Efficiently Scaling Transformer Inference](https://mlsanity.com/paper/2211.05102/) (2022)
- [PyramidKV: Dynamic KV Cache Compression based on Pyramidal Information Funneling](https://mlsanity.com/paper/2406.02069/) (2024)
- [DuoAttention: Efficient Long-Context LLM Inference with Retrieval and Streaming Heads](https://mlsanity.com/paper/2410.10819/) (2024)
- [LLM Inference Unveiled: Survey and Roofline Model Insights](https://mlsanity.com/paper/2402.16363/) (2024)
- [A Survey on Large Language Model Acceleration based on KV Cache Management](https://mlsanity.com/paper/2412.19442/) (2024)
- [PLeak: Prompt Leaking Attacks against Large Language Model Applications](https://mlsanity.com/paper/2405.06823/) (2024)
- [Keep the Cost Down: A Review on Methods to Optimize LLM' s KV-Cache Consumption](https://mlsanity.com/paper/2407.18003/) (2024)
- [A Simple and Effective L2 Norm-Based Strategy for KV Cache Compression](https://mlsanity.com/paper/2406.11430/) (2024)
- [ThinK: Thinner Key Cache by Query-Driven Pruning](https://mlsanity.com/paper/2407.21018/) (2024)
- [Jailbreaking GPT-4V via Self-Adversarial Attacks with System Prompts](https://mlsanity.com/paper/2311.09127/) (2023)
- [KeyDiff: Key Similarity-Based KV Cache Eviction for Long-Context LLM Inference in Resource-Constrained Environments](https://mlsanity.com/paper/2504.15364/) (2025)
- [KV Cache Compression, But What Must We Give in Return? A Comprehensive Benchmark of Long Context Capable Approaches](https://mlsanity.com/paper/2407.01527/) (2024)
- [A Closer Look at System Prompt Robustness](https://mlsanity.com/paper/2502.12197/) (2025)
- [Position is Power: System Prompts as a Mechanism of Bias in Large Language Models (LLMs)](https://mlsanity.com/paper/2505.21091/) (2025)
- [A First Look at GPT Apps: Landscape and Vulnerability](https://mlsanity.com/paper/2402.15105/) (2024)
- [Raccoon: Prompt Extraction Benchmark of LLM-Integrated Applications](https://mlsanity.com/paper/2406.06737/) (2024)
- [Semantic Integrity Matters: Benchmarking and Preserving High-Density Reasoning in KV Cache Compression](https://mlsanity.com/paper/2502.01941/) (2025)
- [LightTransfer: Your Long-Context LLM is Secretly a Hybrid Model with Effortless Adaptation](https://mlsanity.com/paper/2410.13846/) (2024)
- [LagKV: Lag-Relative Information of the KV Cache Tells Which Tokens Are Important](https://mlsanity.com/paper/2504.04704/) (2025)
- [Q-Filters: Leveraging QK Geometry for Efficient KV Cache Compression](https://mlsanity.com/paper/2503.02812/) (2025)

## Built on by (most cited first, 15 of 15)

- [When Does Value-Aware KV Eviction Help? A Fixed-Contract Diagnostic for Non-Monotone Cache Compression](https://mlsanity.com/paper/2605.08234/) (2026)
- [HARD-KV: Head-Adaptive Regularization for Decoding-time KV Compression](https://mlsanity.com/paper/2606.28831/) (2026)
- [Cartridges++: KV Cache Compression without Off-Context Derailment](https://mlsanity.com/paper/2609.35621/) (2026)
- [Free(): Learning to Forget in Malloc-Only Reasoning Models](https://mlsanity.com/paper/2602.08030/) (2026)
- [KV Cache Offloading for Context-Intensive Tasks](https://mlsanity.com/paper/2604.08426/) (2026)
- [DepthWeave-KV: Token-Adaptive Cross-Layer Residual Factorization for Long-Context KV Cache Compression](https://mlsanity.com/paper/2607.06523/) (2026)
- [Adaptive Filtering of the KV Cache: Diagnosing and Correcting Structural-Role Bias in LLM Inference](https://mlsanity.com/paper/2607.13205/) (2026)
- [KVDiagnosis: A Diagnostic Benchmark for KV-Cache Compression in Long-Context Language Models](https://mlsanity.com/paper/2608.09412/) (2026)
- [Random Attention: Rethinking KV Cache Eviction for Efficient Reasoning](https://mlsanity.com/paper/2609.03430/) (2026)
- [FreqDepthKV: Frequency-Guided Depth Sharing for Robust KV Cache Compression in Long-Context LLM Inference](https://mlsanity.com/paper/2607.06519/) (2026)
- [A Probabilistic Interpretation of KV Cache Eviction](https://mlsanity.com/paper/2608.28293/) (2026)
- [DeltaS: Reading the Gated Linear Attention State for KV Cache Eviction in Streaming Video](https://mlsanity.com/paper/2609.27470/) (2026)
- [Risk-Controlled KV-Cache Eviction: From Memory Budgets to Risk Targets](https://mlsanity.com/paper/2609.27981/) (2026)
- [When Keywords Drop but Classifiers Hold: Soft Refusals under KV Cache Compression](https://mlsanity.com/paper/2609.31678/) (2026)
- [To Cache or Not to Cache: Understanding and Mitigating Safety Incompatibility in Cross-Model KV Cache Reuse](https://mlsanity.com/paper/2610.15721/) (2026)

## Abstract (arXiv)

KV cache compression promises increased throughput and efficiency with negligible loss in performance. While the gains in throughput are indisputable and recent literature has indeed shown minimal degradation on particular benchmarks, in general the consequences of compression in realistic scenarios such as multi-instruction prompting have been insufficiently studied. In this paper, we identify several pitfalls that practitioners should be aware of when deploying KV cache compressed LLMs. We evaluate five KV cache compression methods (StreamingLLM, SnapKV, TOVA, H2O, and K-Norm) on Llama3.1 8B and Qwen2.5 14B under multi-instruction prompting with IFEval. Importantly, we show that certain instructions degrade much more rapidly with compression, effectively causing them to be completely ignored by the LLM. As a practical example, we highlight system prompt leakage as a case study, empirically demonstrating the impact of compression on leakage and general instruction-following. We identify several factors that contribute to system prompt leakage: compression method, instruction order, and KV eviction bias. We then propose simple changes to KV cache eviction policies that can reduce the impact of these factors and improve the overall performance in multi-instruction tasks.