# Training Continuous Chain of Thought Models: A Tale of Two Regimes

Authors: Varun Yerram, He He, Eunsol Choi.

Introduces direct latent supervision through averaged trace embeddings and finds that both direct and indirect continuous-CoT training degrade sharply on tasks with substantially longer reasoning traces.

First public: 2026-07-18.
ArXiv initial submission: 2026-07-18.

[Original paper](https://arxiv.org/abs/2607.16972) · [Reading-list entry](https://latent-implicit-thinking.github.io/papers/2607.16972/)
[Preprint BibTeX](https://latent-implicit-thinking.github.io/papers/2607.16972/citation.bib) · [Preprint CSL-JSON](https://latent-implicit-thinking.github.io/papers/2607.16972/citation.csl.json) · [JSON record](https://latent-implicit-thinking.github.io/papers/2607.16972/record.json)
[Code](https://github.com/Varun221/cmtp_research)

## Approaches

- [Continuous-state reasoning](https://latent-implicit-thinking.github.io/topics/continuous-states/): Use vectors—lists of numbers—to carry intermediate reasoning from one step to the next, instead of selecting a text token at every step. Some methods use the model's hidden states; others combine several candidate token representations.
- [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.

Summary and categories are editorial. Read the original paper to verify claims and limitations. A citation to a paper does not, by itself, establish that a later method builds on it.

Reading-list version 0.9.5; literature checked through 2026-10-18.