# MIRAGE: Mobile Agents with Implicit Reasoning and Generative World Models

Authors: Zhichao Yang, Yuanze Hu, Haojie Hao, Longkun Hao, Dongshuo Huang, Hongyu Lin, Gen Li, Lanqing Hong, Yihang Lou, Yan Bai.

Replaces a mobile agent’s written observations, rationale and predicted next screen with continuous thought slots. Parallel refinement updates these slots, while next-frame feature prediction provides training supervision before action decoding.

First public: 2026-06-03.
ArXiv initial submission: 2026-06-03.

[Original paper](https://arxiv.org/abs/2606.04627) · [PDF](https://arxiv.org/pdf/2606.04627) · [HTML](https://arxiv.org/html/2606.04627v2) · [Reading-list entry](https://latent-implicit-thinking.github.io/papers/2606.04627/)
[Preprint BibTeX](https://latent-implicit-thinking.github.io/papers/2606.04627/citation.bib) · [Preprint CSL-JSON](https://latent-implicit-thinking.github.io/papers/2606.04627/citation.csl.json) · [JSON record](https://latent-implicit-thinking.github.io/papers/2606.04627/record.json)

## Approaches

- [Internalized reasoning](https://latent-implicit-thinking.github.io/topics/internalized-reasoning/): Learn from worked solutions so that some or all of the reasoning can later happen without writing those steps. Training may use a separate teacher model or progressively remove written steps from the model being trained.
- [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.

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