# DTCA: Denoising Token Credit Assignment for Reinforcement Learning in Diffusion Language Models

Authors: Hongze Chen, Miao Pan, Lizhou Liang, Qinfeng Li, Xinyin Ma, Yongliang Shen, Weishi Wang, Daniel Dahlmeier, Xuhong Zhang, Wenqi Zhang.

Estimates a token’s contribution by how much revealing it improves predictions at other masked positions. Uses part-of-speech groups to select denoising states and assign reinforcement-learning credit, improving a diffusion language model on math problems.

First public: 2026-10-26.
ArXiv initial submission: 2026-10-26.

[Original paper](https://arxiv.org/abs/2610.35201) · [PDF](https://arxiv.org/pdf/2610.35201) · [HTML](https://arxiv.org/html/2610.35201v1) · [Reading-list entry](https://latent-implicit-thinking.github.io/papers/2610.35201/)
[Preprint BibTeX](https://latent-implicit-thinking.github.io/papers/2610.35201/citation.bib) · [Preprint CSL-JSON](https://latent-implicit-thinking.github.io/papers/2610.35201/citation.csl.json) · [JSON record](https://latent-implicit-thinking.github.io/papers/2610.35201/record.json)

## Approaches

- [Alternative generation methods](https://latent-implicit-thinking.github.io/topics/alternative-reasoning-schedules/): Change how intermediate steps are generated, revised, or searched—for example, by refining several steps in parallel. Some produce written chains of thought.

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