Skip to content
AI News HubLIVE
Original source2 min read

MoEZip: Routing-Aware KV Cache Compression for Sparse Mixture-of-Experts LLMs

Summary

This NeurIPS 2026 paper proposes MoEZip, a KV cache compression method built specifically for sparse Mixture-of-Experts (MoE) language models. Existing compression techniques work well for dense transformers but overlook a problem unique to MoE: evicting KV entries can change which experts are selected when generating an answer, degrading output quality. MoEZip introduces a metric based on the Fisher information matrix of routing distributions to measure how sensitive expert routing is to perturbation, combines it with attention weights for routing-aware scoring, and composes complementary eviction strategies for dense and sparse-MoE architectures. Under aggressive compression with only 10% KV retention, it beats the state-of-the-art baseline by +24 and +20 points on average on LongBench…

SourcearXiv Machine LearningAuthor: Minsang Kim, Seungjun Baek
MoEZip: Routing-Aware KV Cache Compression for Sparse Mixture-of-Experts LLMs
Report an error

The correction channel is not available yet. You can copy the article reference below for later.

Correction instructions
Read article

[Submitted on 21 Oct 2026]

Title:MoEZip: Routing-Aware KV Cache Compression for Sparse Mixture-of-Experts LLMs

View a PDF of the paper titled MoEZip: Routing-Aware KV Cache Compression for Sparse Mixture-of-Experts LLMs, by Minsang Kim and 1 other authors

View PDF HTML (experimental)

Abstract:Mixture-of-Experts (MoE) language models scale model capacity efficiently by activating a small subset of parameters per token. However, in long-context inference, memory remains a major bottleneck because the KV cache grows linearly with sequence length. Existing KV cache compression methods reduce this cost for dense transformers, but overlook a problem specific to MoE: KV eviction can change which experts are selected during answer generation, degrading the generation quality. We propose MoEZip, a routing-aware KV cache compression for sparse MoE language models. To estimate the sensitivity of expert routing to perturbation, we propose a novel metric based on the Fisher information matrix of routing distributions. MoEZip combines this metric with attention weights for routing-aware scoring of contexts. Finally, we compose a complementary scoring system that combines eviction methods for dense and sparse-MoE architectures. Experiments show that MoEZip achieves robust performance under aggressive cache compression on various benchmarks. At a KV retention ratio of 10\%, MoEZip improves over the SoTA baseline by +24 and +20 points on average on LongBench and RULER, corresponding to 2.3$\times$ and 4.9$\times$ higher scores, respectively.

Comments: Annual Conference on Neural Information Processing Systems (NeurIPS) 2026, Accepted

Subjects:

Machine Learning (cs.LG); Artificial Intelligence (cs.AI)

Cite as: arXiv:2610.30206 [cs.LG]

(or arXiv:2610.30206v1 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2610.30206

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Minsang Kim [view email] [v1] Wed, 21 Oct 2026 13:06:30 UTC (5,660 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled MoEZip: Routing-Aware KV Cache Compression for Sparse Mixture-of-Experts LLMs, by Minsang Kim and 1 other authors

View PDF

HTML (experimental)

TeX Source

view license

Additional Features

Audio Summary

Current browse context:

cs.LG

new | recent | 2026-10

Change to browse by:

cs cs.AI

References & Citations

SciX

Google Scholar

Semantic Scholar

Loading...

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

IArxiv recommender toggle

IArxiv Recommender (What is IArxiv?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Key points and analysis

Article intelligence

EngineersAdvanced

Key points

  • Identifies an overlooked MoE-specific problem: KV eviction changes expert selection and degrades generation quality
  • Proposes a routing-sensitivity metric based on the Fisher information matrix of routing distributions, combined with attention weights for scoring
  • At 10% KV retention, improves over the SoTA baseline by +24 and +20 points on average on LongBench and RULER
  • Accepted at NeurIPS 2026; arXiv:2610.30206

Highlights and analysis are generated automatically and may contain errors. Check the original source.

More recent stories on this topic · Models