OSWAM: An Effective World Action Model with One-Step Action Generation

Guowen Zhang; Liyi Chen; Xu Hu; Pengfei Wang; Jingfan Chen; Chenhang He; Lei Zhang

cs.CVOct 26, 2026

OSWAM: An Effective World Action Model with One-Step Action Generation

Authors: Guowen Zhang, Liyi Chen, Xu Hu, Pengfei Wang, Jingfan Chen, Chenhang He, Lei Zhang

Organizations: The Hong Kong Polytechnic University

Abstract

Direct-policy World Action Models (WAM) have demonstrated the benefits of decoupling action generation from explicit future visual prediction for efficient inference in autonomous driving and embodied control. However, they still require tens of denoising steps to achieve competitive action quality, incurring substantial latency and hindering real-time deployment. Simply reducing the number of denoising steps will collapse the nonlinear multi-step denoising trajectory into coarse updates, substantially degrading action quality. To reduce the denoising steps while enabling real-time action generation, we present an effective world action model with one-step action generation, namely OSWAM, for autonomous driving and embodied control. We argue that it is the absence of structured action priors in random noise initialization that forces direct-policy WAMs to recover coherent action sequences through iterative denoising. Hence, we construct imagination tokens to initialize the diffusion process, providing physical priors and action-conditioned representations. We further introduce privileged information distillation with future conditioning to enhance future-aware action generation. As a result, OSWAM can efficiently and effectively generate high quality actions in just one diffusion step. Our experiments demonstrate that OSWAM achieves comparable or even better results in both autonomous driving and embodied control than direct-policy WAM and Joint-WAM that require tens of steps. It obtains approximately $3\times$ speedup over direct-policy WAM baselines and supports real-time compiled inference with latency as low as 40ms on an NVIDIA RTX 5090 GPU. The project page is available at https://gwenzhang.github.io/OSWAM/

Figures & tables

Figure 1: NAVSIM and Robotwin 2.0 inference efficiency-performance comparison. EPT denotes embodied pretraining. All latencies are measured on NVIDIA H20 GPU.

Figure 2: Quantitative comparison of the Fast-WAM [ 4 ] test-time scaling with different action denoising steps on (a) the autonomous driving dataset Navsim, and (b) embodied control dataset RoboTwin.

Figure 3: The overall architecture of OSWAM. A single Video DiT pass produces an observation-conditioned action prior with AIT and complementary scene context with SIT, providing a structured initialization for one-step Action DiT denoising. Privileged-information distillation injects future-informed representations during training, while inference uses only the current observation.

Figure 4: (a) The attention mask of OSWAM and the privileged teacher. (b) Role-aware RoPE explicitly disentangles token groups using unique positional encodings. $sI$ is SIT and $aI$ is AIT.

Method	$N^{v}$	$N^{a}$	NC $\uparrow$	DAC $\uparrow$	EP $\uparrow$	TTC $\uparrow$	C $\uparrow$	PDMS $\uparrow$	Latency $\downarrow$
Human Agent	–	–	100.0	100.0	87.5	100.0	99.9	94.8	–
Traditional E2E and VLA-based planners
UniAD [ 36 ]	–	–	97.8	91.9	78.8	92.9	100.0	83.4	–
TransFuser [ 26 ]	–	–	97.7	92.8	79.2	92.8	100.0	84.0	–
DiffusionDrive [ 8 ]	–	–	98.2	96.2	82.2	94.7	100.0	88.1	–
WoTE [ 31 ]	–	–	98.5	96.8	81.9	94.9	99.9	88.3	–

Table 1: Comparison with state-of-the-art planners on the NAVSIM navtest .

N^{v}

and

N^{a}

denote the number of video and action denoising steps . † SimWAM is reported without reinforcement learning. ‡ DriveVA is reported without CARLA mix training. ∗ Latency reproduced by us.

Method	$N^{v}$	$N^{a}$	EPT	Clean	Rand.	Average	Latency (ms)
VLA-based methods
$\pi_{0}$ [ 5 ]	–	–	w/	65.92	58.40	62.20	–
$\pi_{0.5}$ [ 6 ]	–	–	w/	82.74	76.76	79.80	–
X-VLA [ 45 ]	–	–	w/	72.90	72.80	72.80	–
Imagine-then-execute-based WAM methods
Motus [ 16 ]	–	–	w/	88.66	87.02	87.80	2367

Table 2: Success rates on RoboTwin 2.0 simulation (Clean and Randomized splits, 50 tasks). EPT denotes embodied pretraining.

Dataset	Model	VAE GPU Time	Visual GPU Time	Action GPU Time	Peak Mem. (GB)	Overall Latency	Compiled Latency
RoboTwin 2.0	LingBot-VA	32.65	1165.12	1205.38	19.12	3365.35	–
	Motus	11.14	567.37	Shared	17.95	1472.85	–
	Fast-WAM	5.27	18.45	67.61	12.70	323.67	73.64
	OSWAM	5.27	23.93	6.78	12.75	95.35	40.81
Navsim	SimWAM	9.83	22.80	72.78	12.87	330.8	90.53
Navsim	OSWAM	9.82	30.87	7.32	12.90	99.70	52.64

Table 3: Inference efficiency on RoboTwin 2.0 and Navsim. Latency is measured with cached language context on a single NVIDIA RTX 5090 GPU and Intel Xeon Platinum 8470Q. All time are reported in milliseconds (ms).

Table 8

Table 9

	Fast-WAM		OSWAM (Ours)
Task	SR $\uparrow$	Time (s) $\downarrow$	SR $\uparrow$	Time (s) $\downarrow$
Bread into bowl	80	$24.6$	85	$\mathbf{16.2}$
Eraser into drawer	75	$17.6$	80	$\mathbf{13.8}$

Table 8: Real-world evaluation on PiPER. SR: success rate (%). Time: video-estimated completion time over successful episodes.

Figure 5: Real-world manipulation with OSWAM on a PiPER robotic arm: (a) placing bread into a bowl and (b) placing a board eraser into an open drawer.

Appendix figures & tables4 assets

Supplementary material from the paper’s appendix.

Appendix

Traditional E2E and VLA-based planners
Method	$N^{v}$	$N^{a}$	NC $\uparrow$	DAC $\uparrow$	DDC $\uparrow$	TLC $\uparrow$	EP $\uparrow$	TTC $\uparrow$	LK $\uparrow$	HC $\uparrow$	EC $\uparrow$	EPDMS $\uparrow$
Human Agent	–	–	100.0	100.0	99.8	100.0	87.4	100.0	100.0	98.1	90.1	90.3
TransFuser [ 26 ]	–	–	96.9	89.9	97.8	99.7	87.1	95.4	92.7	98.3	87.2	76.7
Hydra-MDP++ [ 49 ]	–	–	97.2	97.5	99.4	99.6	83.1	96.5	94.4	98.2	70.9	81.4
GTRS-Dense [ 50 ]	–	–	97.6	97.5	99.0	99.9	87.9	97.0	95.9	97.5	55.9	82.3
DriveSuprim [ 51 ]	–	–	97.5	96.5	99.4	99.6	88.4	96.6	95.5	98.3	77.0	83.1

Appendix

Table A.1: Comparison with state-of-the-art planners on the NAVSIM-v2 navtest .

N^{v}

and

N^{a}

denote the number of video and action denoising steps, respectively.

Method	Auxiliary Supervision	L2 (m) $\downarrow$				Collision Rate (%) $\downarrow$
Method	Auxiliary Supervision	1s	2s	3s	Avg.	1s	2s	3s	Avg.
ST-P3 [ 55 ]	Map & Box & Depth	1.33	2.11	2.90	2.11	0.23	0.62	1.27	0.71
UniAD [ 36 ]	Map & Box & Motion & Tracklets & Occ	0.48	0.96	1.65	1.03	0.05	0.17	0.71	0.31
OccNet [ 56 ]	3D-Occ & Map & Box	1.29	2.13	2.99	2.14	0.21	0.59	1.37	0.72
OccWorld [ 30 ]	3D-Occ	0.52	1.27	2.41	1.40	0.12	0.40	2.08	0.87
VAD-Tiny [ 57 ]	Map & Box & Motion	0.60	1.23	2.06	1.30	0.31	0.53	1.33	0.72

Appendix

Table A.2: End-to-end motion planning performance on the nuScenes [ 47 ] dataset.

Table 14

Figure B.1: Qualitative comparison of one-step SimWAM and OSWAM on Navsim v1.

Efficient Inference for World Action Models

Efficient inference for world action models cuts the cost of predicting action-conditioned futures for responsive control and planning.

See all 80 papers

1Step-WAM: Your World Action Model but Faster using 1-Step Action Generation
Oct 11, 2026Rundong Su, Pengtao Chen, Haohan Geng +6Efficient Inference for World Action Models World Action Models
Fudan University · University of Science and Technology of China · Peking University +1
SimWAM: A Simple World Action Model for End-to-End Autonomous Driving
Aug 7, 2026Zongchuang Zhao, Xin Zhou, Tianyang Xu +6End-to-End Autonomous Driving Autonomous Driving
Huazhong University of Science & Technology · Dongfeng Research & Development Institute
The DAWN of World-Action Interactive Models
May 12, 2026Hongbo Lu, Liang Yao, Chenghao He +6Diffusion-Based Planning World Model-Based Planning
1COWARobot Co. Ltd · 2Shanghai Jiao Tong University · 3Hohai University

Oct 11, 2026J/K move · Enter open · S save

1Step-WAM: Your World Action Model but Faster using 1-Step Action Generation

Rundong Su, Pengtao Chen, Haohan Geng, Sha Zhang+5

Fudan University · University of Science and Technology of China · Peking University+1

Efficient Inference for World Action Models World Action Models One-Step Generative Modeling

Open paper AskPDF arXiv

World Action Models (WAMs) unify visual world modeling and action generation, yet their real-world deployment is heavily bottlenecked by iterative action denoising and the sequential execution of visual prefilling and action prediction. We demonstrate that the action-denoising trajectories in WAMs are approximately straight, rendering the first-step velocity a strong estimate of the final action direction. Building on this observation, we introduce 1Step-WAM, a framework that transforms multi-step action denoising into one single forward. Specifically, we employ a lightweight calibrator to directly correct the first step velocity, and use bandwidth-constrained action decoding to further reduce errors in the calibrated one-step actions. Compressing action generation into a single step enables it to execute in parallel with visual prefilling, completely breaking the sequential bottleneck of the traditional imagination-then-action paradigm. Evaluations on RoboTwin and LIBERO show that 1Step-WAM largely preserves multi-step task success with only about 13.4M trainable parameters, while achieving inference speedups of

7.44\times

5.98\times

, and

24.84\times

for FastWAM, ImageWAM and Kairos, respectively. Furthermore, real-world robotic deployments using the FastWAM backbone confirm a consistent 7.44

\times

speedup while maintaining close success rate to the multi-step baseline. Code is avaliable at https://github.com/changdong1687/1Step-WAM.

Abstract

Efficient Inference for World Action Models

Explore similar work

1Step-WAM: Your World Action Model but Faster using 1-Step Action Generation

SimWAM: A Simple World Action Model for End-to-End Autonomous Driving

The DAWN of World-Action Interactive Models