# Recency Is the Wrong Axis: A Negative Result on Windowed Deepest-Plane KV Caching in Looped Language Models

Authors: Dezhen Lu, Runguo Li, Zixiang Ni, Yuchen Tang, Xingquan Kong, Jiaxing Yang, Jiarui Guo.

Tests a cache policy that retains every loop’s keys and values for recent positions and shares the deepest loop’s cache elsewhere. Across three looped language models, recent-position windows fail to provide a reliable repair; convergence across loops and which positions share matter more.

First public: 2026-10-25.
ArXiv initial submission: 2026-10-25.

[Original paper](https://arxiv.org/abs/2610.34436) · [PDF](https://arxiv.org/pdf/2610.34436) · [HTML](https://arxiv.org/html/2610.34436v2) · [Reading-list entry](https://latent-implicit-thinking.github.io/papers/2610.34436/)
[Preprint BibTeX](https://latent-implicit-thinking.github.io/papers/2610.34436/citation.bib) · [Preprint CSL-JSON](https://latent-implicit-thinking.github.io/papers/2610.34436/citation.csl.json) · [JSON record](https://latent-implicit-thinking.github.io/papers/2610.34436/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.
- [Analysis and evaluation](https://latent-implicit-thinking.github.io/topics/implicit-reasoning-evaluation/): Measure how well these methods work, investigate what their internal computation does, and identify where they fail. This includes experiments, studies of model internals, and mathematical results.

Reading-list version 0.12.0; literature checked through 2026-11-02.