# LoopKV: Loop-Aware KV Cache Compression for Serving Looped Transformers

Authors: Tianle Zhou, Yang Zhang, Tianyu Wu, Zhuohan Wang, William Nixon, Juncheng Yang.

Allocates cache precision separately for each loop and layer, shares selected caches across loops, and reads low-attention values at lower precision to reduce the memory and latency of looped language models.

First public: 2026-11-03.
ArXiv initial submission: 2026-11-03.

[Original paper](https://arxiv.org/abs/2611.04137) · [PDF](https://arxiv.org/pdf/2611.04137) · [HTML](https://arxiv.org/html/2611.04137v1) · [Reading-list entry](https://latent-implicit-thinking.github.io/papers/2611.04137/)
[Preprint BibTeX](https://latent-implicit-thinking.github.io/papers/2611.04137/citation.bib) · [Preprint CSL-JSON](https://latent-implicit-thinking.github.io/papers/2611.04137/citation.csl.json) · [JSON record](https://latent-implicit-thinking.github.io/papers/2611.04137/record.json)

## Approaches

- [Recurrent reasoning](https://latent-implicit-thinking.github.io/topics/recurrent-reasoning/): Repeat parts of the model or keep intermediate information in memory while working through a problem. Some methods also learn how many passes to run.

Reading-list version 0.15.2; literature checked through 2026-11-03.