# Agreement rather than normalisation keeps depth reusable in looped language models

Authors: Ivan Viakhirev, Kirill Borodin, Maxim Abramov, Grach Mkrtchian.

Tests normalization between passes through a looped transformer. Removing it from pretrained Ouro models breaks predictions, while small models trained without it generalize better across loop counts, showing why the computation used during training and inference must agree.

First public: 2026-10-22.
ArXiv initial submission: 2026-10-22.

[Original paper](https://arxiv.org/abs/2610.31943) · [PDF](https://arxiv.org/pdf/2610.31943) · [HTML](https://arxiv.org/html/2610.31943v1) · [Reading-list entry](https://latent-implicit-thinking.github.io/papers/2610.31943/)
[Preprint BibTeX](https://latent-implicit-thinking.github.io/papers/2610.31943/citation.bib) · [Preprint CSL-JSON](https://latent-implicit-thinking.github.io/papers/2610.31943/citation.csl.json) · [JSON record](https://latent-implicit-thinking.github.io/papers/2610.31943/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 0.11.0; literature checked through 2026-10-26.