# How Depth Growth and Looping Shape In-Context Learning in Language Models

Authors: Ferdinand Kapl, Kaitlin Maile, Stefan Bauer, Johannes von Oswald.

Compares dense, progressively grown and looped Transformers on in-context learning. Attention-head tracing and ablation link stronger context use in grown and looped models to greater reliance on induction heads.

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

[Original paper](https://arxiv.org/abs/2610.31421) · [PDF](https://arxiv.org/pdf/2610.31421) · [HTML](https://arxiv.org/html/2610.31421v1) · [Reading-list entry](https://latent-implicit-thinking.github.io/papers/2610.31421/)
[Preprint BibTeX](https://latent-implicit-thinking.github.io/papers/2610.31421/citation.bib) · [Preprint CSL-JSON](https://latent-implicit-thinking.github.io/papers/2610.31421/citation.csl.json) · [JSON record](https://latent-implicit-thinking.github.io/papers/2610.31421/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.