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

arXiv: https://arxiv.org/abs/2606.04627 · HTML: https://mlsanity.com/paper/2606.04627/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in World models for language agents, one of 11 directions in World models.
- Builds on 22 of its 32 references that are in the map.
- First posted on arXiv 3 Jun 2026.

## Placed in

- World models / [World models for language agents](https://mlsanity.com/wm/world-models-for-language-agents/)
- LLM agents and tool use / [GUI and mobile agents](https://mlsanity.com/agents/gui-and-mobile-agents/)

## Builds on (most cited first, 22 of 22)

- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [GPT-4o System Card](https://mlsanity.com/paper/2410.21276/) (2024)
- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [Dream to Control: Learning Behaviors by Latent Imagination](https://mlsanity.com/paper/1912.01603/) (2019)
- [World Models](https://mlsanity.com/paper/1803.10122/) (2018)
- [Training Large Language Models to Reason in a Continuous Latent Space](https://mlsanity.com/paper/2412.06769/) (2024)
- [UI-TARS: Pioneering Automated GUI Interaction with Native Agents](https://mlsanity.com/paper/2501.12326/) (2025)
- [Revisiting Feature Prediction for Learning Visual Representations from Video](https://mlsanity.com/paper/2404.08471/) (2024)
- [AndroidWorld: A Dynamic Benchmarking Environment for Autonomous Agents](https://mlsanity.com/paper/2405.14573/) (2024)
- [OS-ATLAS: A Foundation Action Model for Generalist GUI Agents](https://mlsanity.com/paper/2410.23218/) (2024)
- [Fast-WAM: Do World Action Models Need Test-time Future Imagination?](https://mlsanity.com/paper/2603.16666/) (2026)
- [Quiet-STaR: Language Models Can Teach Themselves to Think Before Speaking](https://mlsanity.com/paper/2403.09629/) (2024)
- [Mobile-Agent: Autonomous Multi-Modal Mobile Device Agent with Visual Perception](https://mlsanity.com/paper/2401.16158/) (2024)
- [Think before you speak: Training Language Models With Pause Tokens](https://mlsanity.com/paper/2310.02226/) (2023)
- [GUI-R1 : A Generalist R1-Style Vision-Language Action Model For GUI Agents](https://mlsanity.com/paper/2504.10458/) (2025)
- [Implicit Chain of Thought Reasoning via Knowledge Distillation](https://mlsanity.com/paper/2311.01460/) (2023)
- [UI-Venus Technical Report: Building High-performance UI Agents with RFT](https://mlsanity.com/paper/2508.10833/) (2025)
- [MAI-UI Technical Report: Real-World Centric Foundation GUI Agents](https://mlsanity.com/paper/2512.22047/) (2025)
- [ViMo: A Generative Visual GUI World Model for App Agents](https://mlsanity.com/paper/2504.13936/) (2025)
- [MobileDreamer: Generative Sketch World Model for GUI Agent](https://mlsanity.com/paper/2601.04035/) (2026)
- [Ferret-UI Lite: Lessons from Building Small On-Device GUI Agents](https://mlsanity.com/paper/2509.26539/) (2025)
- [TinyAlign: Boosting Lightweight Vision-Language Models by Mitigating Modal Alignment Bottlenecks](https://mlsanity.com/paper/2505.12884/) (2025)

## Built on by (most cited first, 1 of 1)

- [Right Screen, Wrong Transition: World Models as Verifiers for GUI Agents](https://mlsanity.com/paper/2610.11942/) (2026)

## Abstract (arXiv)

Mobile agents are increasingly expected to operate everyday applications from screenshots and language goals, where reliable control requires reasoning over screen affordances, multi-step navigation, and future state changes. However, many agents externalize this computation as long textual chains of thought, which slows interaction, increases supervision cost, and complicates deployment. We introduce MIRAGE, a framework that learns continuous latent reasoning representations from visible textual reasoning traces. MIRAGE transfers explicit reasoning into compact hidden states, enabling the agent to reason internally without decoding long rationales. It also incorporates a generative world-model objective: latent reasoning vectors are aligned with future screenshots, encouraging the agent to anticipate upcoming interface states before acting. This turns hidden computation into both a compressed thought representation and a forward-looking model of environment dynamics. At inference time, MIRAGE reasons in continuous latent space, reducing token generation while improving execution efficiency. On AndroidWorld, MIRAGE matches explicit chain-of-thought supervised fine-tuning in the 4B ablation with a 3-5x lower decoded-token budget and improves a comparable instruction-tuned baseline by 10.2 points; on AndroidControl, it improves action grounding while generating over 75% fewer tokens.