# Agile-WAM: An Agile Tactile World Action Model for Contact-Rich Robot Control

arXiv: https://arxiv.org/abs/2609.20761 · HTML: https://mlsanity.com/paper/2609.20761/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Embodied and robot world models, one of 11 directions in World models.
- Builds on 21 of its 21 references that are in the map.
- First posted on arXiv 17 Sep 2026.

## Placed in

- World models / [Embodied and robot world models](https://mlsanity.com/wm/embodied-and-robot-world-models/)

## Builds on (most cited first, 21 of 21)

- [Flow Matching for Generative Modeling](https://mlsanity.com/paper/2210.02747/) (2022)
- [π0: A Vision-Language-Action Flow Model for General Robot Control](https://mlsanity.com/paper/2410.24164/) (2024)
- [World Action Models are Zero-shot Policies](https://mlsanity.com/paper/2602.15922/) (2026)
- [Unified Video Action Model](https://mlsanity.com/paper/2503.00200/) (2025)
- [Reactive Diffusion Policy: Slow-Fast Visual-Tactile Policy Learning for Contact-Rich Manipulation](https://mlsanity.com/paper/2503.02881/) (2025)
- [FlowPolicy: Enabling Fast and Robust 3D Flow-based Policy via Consistency Flow Matching for Robot Manipulation](https://mlsanity.com/paper/2412.04987/) (2024)
- [Prediction with Action: Visual Policy Learning via Joint Denoising Process](https://mlsanity.com/paper/2411.18179/) (2024)
- [Learning Robotic Manipulation Policies from Point Clouds with Conditional Flow Matching](https://mlsanity.com/paper/2409.07343/) (2024)
- [Affordance-based Robot Manipulation with Flow Matching](https://mlsanity.com/paper/2409.01083/) (2024)
- [AdaFlow: Imitation Learning with Variance-Adaptive Flow-Based Policies](https://mlsanity.com/paper/2402.04292/) (2024)
- [AnyTouch: Learning Unified Static-Dynamic Representation across Multiple Visuo-tactile Sensors](https://mlsanity.com/paper/2502.12191/) (2025)
- [FACTR: Force-Attending Curriculum Training for Contact-Rich Policy Learning](https://mlsanity.com/paper/2502.17432/) (2025)
- [VITA: Vision-to-Action Flow Matching Policy](https://mlsanity.com/paper/2507.13231/) (2025)
- [Dream-Tac: A Unified Tactile World Action Model for Contact-Rich Robot Manipulation](https://mlsanity.com/paper/2606.08737/) (2026)
- [VTAM: Video-Tactile-Action Models for Complex Physical Interaction Beyond VLAs](https://mlsanity.com/paper/2603.23481/) (2026)
- [ManiFeel: Benchmarking and Understanding Visuotactile Manipulation Policy Learning](https://mlsanity.com/paper/2505.18472/) (2025)
- [AnyTouch 2: General Optical Tactile Representation Learning For Dynamic Tactile Perception](https://mlsanity.com/paper/2602.09617/) (2026)
- [Tactile-WAM: Touch-Aware World Action Model with Tactile Asymmetric Attention](https://mlsanity.com/paper/2606.26663/) (2026)
- [Sensor-Invariant Tactile Representation](https://mlsanity.com/paper/2502.19638/) (2025)
- [IN-RIL: Interleaved Reinforcement and Imitation Learning for Policy Fine-Tuning](https://mlsanity.com/paper/2505.10442/) (2025)
- [WorldAgen: Unified State-Action Prediction with Test-Time World Model Training](https://mlsanity.com/paper/2609.08162/) (2026)

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

- [DuoMind: Enabling Distributed Multi-Robot Coordination with Semantic Communication](https://mlsanity.com/paper/2610.02161/) (2026)

## Abstract (arXiv)

World Action Models (WAMs) advance beyond conventional visuomotor policies by jointly predicting future world states and robot actions, enabling the policy to learn phys- ical dynamics that support effective control. However, recent tactile WAMs often rely on large-scale pretrained generative backbones to capture contact-rich physical dynamics, which limit their inference efficiency and flexible deployment. In this paper, we present Agile-WAM, an agile tactile World Action Model for contact-rich robot control. Agile-WAM encodes visual and tactile observations into a shared latent that serves as the source of a direct vision-tactile-to-action flow-matching process, which can jointly generate latent representations of action chunks and future visual/tactile latents. A key observation is that vision and tactile signals evolve at inherently different timescales: adjacent visual frames are often highly similar, whereas tactile signals can change abruptly upon contact. We therefore introduce multi-horizon multimodal prediction in Agile-WAM, which provides supervision for visual latent at a larger temporal offset while predicting the tactile latent in the next frame to capture fine-grained contact dynamics. Across nine simulated and five real-world contact-rich ma- nipulation tasks, Agile-WAM demonstrates strong and robust performance, outperforming the strongest baseline in success rate while maintaining low inference latency. In particular, in five real-world experiments, Agile-WAM yields a relative gain of 29.4% in overall success rates while achieving inference latency of 11.9 ms. These results demonstrate that multimodal WAM can be achieved with an agile architecture suitable for precise and high-frequency robot control. More details are available on our project page.