| title | RealtimeWAM: One-Step Asynchronous World Action Models |
|---|---|
| authors | Chengtao Lv, Jinyang Du, Shuyi Feng, Yang Yong, Shiqiao Gu, Shunzi Yang, Ruihao Gong, Shen Ren, Tianwei Zhang, Wenya Wang |
| date | 2026-10-05 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2610.06617 |
| affiliation | Nanyang Technological University |
| upvotes | 12 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf code |
| abstract | World Action Models (WAMs) incorporate visual representations from video generation backbones to guide action prediction. Recent efficient WAMs adopt Mixture-of-Transformers (MoT) architectures and compute video representations once for reuse by the action expert. However, intra-expert iteration (\ie, multi-step action denoising) and inter-expert waiting (\ie, sequential execution of the video and action experts) still limit inference efficiency. To this end, we present RealtimeWAM, an extremely efficient WAM variant with one-step action generation and asynchronous inference, addressing these two bottlenecks. To reduce intra-expert iteration, we propose Teacher-Anchored Consistency Distillation (TACD) to address a local-global error gap: low local consistency error alone does not guarantee accurate final actions. TACD supplements local consistency with explicit supervision from the frozen teacher's multi-step rollout endpoint, enabling accurate one-step action generation. Additionally, we propose Cross-Expert Wavefront Pipelining (CEWP) to eliminate unnecessary expert-level waiting. It overlaps the two experts through block-wise sharing of the video KV cache, synchronizing only immediately before the corresponding action attention consumes it. Extensive experiments across diverse benchmarks (\eg, LIBERO, LIBERO-Plus and RoboTwin) and model variants (\eg, Fast-WAM and Faster-WAM) demonstrate the superiority of RealtimeWAM. Notably, RealtimeWAM maintains near-lossless performance (\ie, <1% drop) across these benchmarks while delivering significant end-to-end speedup (\eg, sim25times on H100). Our code and checkpoints are available via this https://github.com/ModelTC/LightX2V/tree/main/examples/realtimewam{link}. |
cite
@misc{lv2026realtimewam,
title={RealtimeWAM: One-Step Asynchronous World Action Models},
author={Chengtao Lv and Jinyang Du and Shuyi Feng and Yang Yong and Shiqiao Gu and Shunzi Yang and Ruihao Gong and Shen Ren and Tianwei Zhang and Wenya Wang},
year={2026},
eprint={2610.06617},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2610.06617}
}