Do World Action Models Generalize Better than VLAs? A Robustness Study

Zhanguang Zhang; Zhiyuan Li; Behnam Rahmati; Rui Heng Yang; Yintao Ma; Amir Rasouli; Sajjad Pakdamansavoji; Yangzheng Wu

cs.ROMar 23, 2026

Do World Action Models Generalize Better than VLAs? A Robustness Study

Authors: Zhanguang Zhang, Zhiyuan Li, Behnam Rahmati, Rui Heng Yang, Yintao Ma, Amir Rasouli, Sajjad Pakdamansavoji, Yangzheng Wu, +6 more

Organizations: Huawei Technologies · University of Toronto

Abstract

Robot action planning in the real world is challenging as it requires not only understanding the current state of the environment but also predicting how it will evolve in response to actions. Vision-language-action (VLA), which repurpose large-scale vision-language models for robot action generation using action experts, have achieved notable success across a variety of robotic tasks. Nevertheless, their performance remains constrained by the scope of their training data, exhibiting limited generalization to unseen scenarios and vulnerability to diverse contextual perturbations. More recently, world models have been revisited as an alternative to VLAs. These models, referred to as world action models (WAMs), are built upon world models that are trained on large corpora of video data to predict future states. With minor adaptations, their latent representation can be decoded into robot actions. It has been suggested that their explicit dynamic prediction capacity, combined with spatiotemporal priors acquired from web-scale video pretraining, enables WAMs to generalize more effectively than VLAs. In this paper, we conduct a comparative study of prominent state-of-the-art VLA policies and recently released WAMs. We evaluate their performance on the LIBERO-Plus and RoboTwin 2.0-Plus benchmarks under various visual and language perturbations. Our results show that WAMs achieve strong robustness, with LingBot-VA reaching 74.2% success rate on RoboTwin 2.0-Plus and Cosmos-Policy achieving 82.2% on LIBERO-Plus. While VLAs such as $π_{0.5}$ can achieve comparable robustness on certain tasks, they typically require extensive training with diverse robotic datasets and varied learning objectives. The evaluation code for the RoboTwin2.0-Plus benchmark is available at: https://robot-robustness.github.io/RoboTwin2.0-Plus/.

Figures & tables

Model	#Params	Backbone	MOT	Pretrain Free	Causal Pred.	AR Gen.
VPP ( Hu et al., 2025 )	1.5B	Stable Video Diffusion (SVD)	✗	✗	✓	✗
GE-Act ( Liao et al., 2025 )	2.2B	LTX-Video-2B	✓	✗	✗	✓
Cosmos-Policy ( Kim et al., 2026 )	2B	Cosmos-Predict2-2B	✗	✓	✗	✗
LingBot-VA ( Li et al., 2026b )	5.3B	Wan2.2-5B	✓?	✗	✓	✓
DreamZero ( Ye et al., 2026b )	14B	Wan2.1-14B	✗	✗	✗	✓
GigaWorld-Policy ( Ye et al., 2026a )	>5B	Wan2.2-5B	✗	✗	✓	✗

Table 1: Summary of world action models (WAMs) . MOT : mixture-of-transformers, indicating that separate transformer backbones are used for video and action streams, with cross-modal interactions facilitated via attention at each layer. Note that the released LingBot-VA model adopts a unified transformer for both modalities, instead of the mixture-of-transformer architecture described in the paper. Pretrain Free : the method does not require task-agnostic embodied pre-training. Causal Pred. : action prediction is conditioned on the generated visual state, or vice versa. AR Gen. : auto-regressive generation.

VLAs
Model	Semantic Understanding & Spatial Grounding	General Video	Robot Data	Task-specific Data
$\pi_{0}$	-	-	Cross-embodiment (>10k h)	Trajectories (5–100 h)
OpenVLA-OFT	-	-	Cross-embodiment (970k)	Trajectories (20–300)
X-VLA	-	-	Cross-embodiment (288k)	Trajectories (50)
$\pi_{0.5}$	High-level planning	-	Mobile manipulation (400 h)	Trajectories (1–20 h)
	Web data-VQA		Multi-env. tabletop

Table 2: Training data of VLAs and WAMs across different training stages. Training stages : Embodied Pre-training , Embodied Post-training , Task-specific Finetuning . PT : filtered data for embodied post-training.

Figure 1: Examples of perturbations on RoboTwin 2.0-Plus tasks. Definition of notations (N1, L1, etc.) can be found in Appendix A

Model	Original	Camera	Robot	Lang.	Light	BG	Noise	Layout	Total
VLAs
$\pi_{0.5}$	78.4	45.6	27.6	74.4	49.6	71.7	64.9	56.8	58.6
X-VLA	65.6	23.2	65.2	64.4	63.1	58.6	49.7	34.8	53.1
VLA + WM
MOTUS	87.0	21.6	85.0	83.2	84.6	84.4	43.1	82.8	71.5
WAMs

Table 3: RoboTwin 2.0-Plus evaluation results . The Original column reports the success rate of each method evaluated in the original RoboTwin 2.0 Easy setting. For a fair comparison, all methods use a single unified model across all tasks. The best result in each column is highlighted in bold , while second-best result is underlined . Lang. : Language. BG : Background.

Figure 2: Case studies on RoboTwin 2.0-Plus. We present three representative cases comparing $\pi_{0.5}$ and LingBot-VA. Keyframes are shown sequentially from left to right. (a) Task: beat block with hammer. Perturbation: noise (N3). Result: $\pi_{0.5}$ collides with the hammer and fails to complete the task. (b) Task: handover block. Perturbation: layout. Result: $\pi_{0.5}$ collides with the red block during approach, leading to failure. (c) Task: rank RGB blocks. Perturbation: lighting (L1–L4 mix). Result: $\pi_{0.5}$ fails to grasp the first red block due to misalignment and does not recover. LingBot-VA success in all three cases shown.

Model	Original	Camera	Robot	Lang.	Light	BG	Noise	Layout	Total
VLAs
$\pi_{0}$	94.2	13.8	6.0	58.8	85.0	81.4	79.0	68.9	53.6
$\pi_{0}(\text{rerun})$	91.3	61.0	40.8	63.5	89.3	84.1	80.1	76.4	69.4
$\pi_{0}$ -FAST	85.5	65.1	21.6	61.0	73.2	73.2	74.4	68.8	61.6
$\pi_{0.5}$	96.9	75.4	77.5	85.6	96.9	94.6	89.7	85.7	85.7
Openvla-OFT_m	97.6	55.6	21.7	81.0	92.7	91.0	78.6	68.7	67.9

Table 4: LIBERO-Plus evaluation results . The Original column denotes results on the standard LIBERO benchmark. The results of

\pi_{0}

\pi_{0}\text{-FAST}

, Openvla-OFT-m, UniVLA, RIPT-VLA are sourced from Fei et al. (2025) . The results reported for HoloBrain0-GD ( Lin et al., 2026 ) and ABot-M0 ( Yang et al., 2026 ) are taken from their respective papers. The remaining models are evaluated using their official repositories and publicly released checkpoints. Specifically,

\pi_{0}(\text{rerun})

and

\pi_{0.5}

are evaluated using their JAX implementations. The best result in each column is highlighted in bold , while second-best result is underlined . Lang. : Language. BG : Background.

Figure 3: Illustrations of future images predicted by Cosmos-policy. Ground-truth and predicted images under three types of perturbations (noise, lighting, and background variations) in LIBERO-Plus are shown. GT: the ground truth. Pred.: the predictions generated by Cosmos-policy.

	$\pi_{0.5}$	X-VLA	Fast-WAM ∗	GE-Act	Cosmos-Policy	LingBot-VA(RW)	MOTUS	LingBot-VA(RT)
Action chunk size	50	30	32	36	16	32	16	32
Inference $t$ (wall clock)	63 ms	195 ms	190 ms	300 ms	390 ms	480 ms	1175 ms	5230 ms
Inference $t$ (w.r.t. $\pi_{0.5}$ )	1.0	3.1 $\times$	3.0 $\times$	4.8 $\times$	6.2 $\times$	7.6 $\times$	18.6 $\times$	83.0 $\times$

Table 5: Runtime Speed Comparison. The inference times of all models are evaluated using the same device. LingBot-VA(RW): real-world setting with three denoising steps for state and five for action. LingBot-VA(RT): RoboTwin 2.0 setting with 25 denoising steps for state and 50 for action.

Appendix figures & tables5 assets

Supplementary material from the paper’s appendix.

Appendix

Dimension	Code	Sub-dimension	Implementation Details
Sensor Noise	N1	Motion blur	Gaussian kernel ( $r\in[3,15]$ , $\sigma\in[1,8]$ ), rotated by random angle $\in[-30^{\circ},30^{\circ}]$
	N2	Gaussian blur	Isotropic Gaussian ( $\sigma\in[1,10]$ )
	N3	Zoom blur	Multi-scale warp accumulation ( $s\in[1.0,1.56]$ , step $\in[0.01,0.03]$ )
	N4	Fog	Homogeneous white fog with transmittance $e^{-\alpha d}$ ( $\alpha\in[0.3,1.5]$ , $d{=}3$ )
	N5	Glass blur	Pixel displacement ( $\delta\in[1,5]$ px) + Gaussian ( $\sigma\in[0.5,2.5]$ ), 1–3 iterations
Light Conditions	L1	Diffuse color	Per-channel RGB tint $\in[0.0,3.5]$ applied to all directional and point lights (always on)

Appendix

Table 6: RoboTwin 2.0-Plus perturbation taxonomy. All 21 sub-dimensions across 7 LIBERO-Plus dimensions, with implementation-level parameter details.

#	Config	Perturbation	Active Sub-dims
0	demo_clean	None (clean baseline)	–
1	demo_vision_noise	Sensor Noise	N1–N5 (cycled)
2	demo_light	Lighting	L1 (always) + L2/L3/L4 (stochastic)
3	demo_camera	Camera Viewpoints	C1 + C3 (C2 disabled by default)
4	demo_robot_state	Robot Initial State	Joint noise + gripper extremes
5	demo_background_plus	Background	B1 (texture + color tint) + B2

Appendix

Table 7: RoboTwin-Plus evaluation configs. The 8 required configs for a complete robustness evaluation.

Aspect	LIBERO-Plus	RoboTwin 2.0-Plus
Simulator	MuJoCo (robosuite)	SAPIEN (ManiSkill3)
Robot	Franka Panda (7-DoF)	Aloha-AgileX (14-DoF)
Arms	Single arm	Dual arm (bimanual)
Cameras	2 (third-person + wrist)	3 (head + 2 wrist)
Image resolution	$256\times 256$	$320\times 240$
Native action space	7-dim delta EEF	14-dim joint positions

Appendix

Table 8: Benchmark comparison between LIBERO-Plus and RoboTwin 2.0-Plus.

Model	Action Repr.	Dim	Chunk	Freq
VLA Models
Pi0	Delta EEF	7	50	50 Hz
Pi0-FAST	Delta EEF (FAST tokenized)	7	50	15–50 Hz
Pi0.5	Absolute EEF / joints	18–19	50	50 Hz
OpenVLA-OFT	Delta EEF	7	8	3–10 Hz
UniVLA	Delta EEF (DCT tokenized)	7	10	–

Appendix

Table 9: Model setup comparison. Action representation, dimensionality, chunk size, and control frequency for each evaluated model. Entries marked with “–” indicate values inherited from the base model or not explicitly specified. Unless noted, models on LIBERO follow the native axis-angle rotation convention.

Category	Model	Camera	Robot	Lang.	Light	BG	Noise	Layout	Total
Spatial	$\pi_{0}$	70.7	49.1	67.9	92.8	95.0	87.7	94.0	78.6
	$\pi_{0.5}$	76.9	85.4	90.3	97.6	95.3	91.7	97.7	90.3
	X-VLA	24.5	91.4	82.1	100.0	100.0	63.5	93.8	77.7
	VLA-JEPA	74.2	71.7	96.1	97.9	99.6	77.8	95.3	86.9
	GE-Act	72.3	75.7	90.8	100.0	89.1	97.7	89.6	87.5
	Cosmos-policy	81.1	59.7	88.2	89.0	91.1	98.0	99.2	86.6

Appendix

Table 10: Detailed LIBERO-Plus results for each task suite. Lang. : Language. BG : Background. The Fast-WAM LIBERO checkpoint was trained on clean demonstrations only, whereas its RoboTwin checkpoint was trained on clean + domain-randomized demonstrations (cf. RQ 2 ).

Explore similar work

WAVE-0: Bridging VLA Policies and WAMs with Lightweight Latent Future Prediction
Oct 13, 2026Haomin Wang, Anqiong Ji, Ke Wang +4World Models for Robotics World Action Models
AutoLab, School of Artificial Intelligence, Shanghai Jiao Tong University · Baidu AI Cloud · Awaken Robotics +1
LaWAM: Latent World Action Models for Efficient Dynamics-Aware Robot Policies
Jun 14, 2026Jialei Chen, Kai Wang, Kang Chen +9Latent Action Learning Action-Conditioned World Models
Jilin University · Zhongguancun Academy · Nankai University +4
Think Like a World Model, Act Like a VLA: Distilling World-Model Representations into Compact Robot Policies
Sep 21, 2026Trung Dao, Sankalp Yamsani, Jaden Park +2World Models Robotic Manipulation
University of Wisconsin–Madison, USA. · University of Illinois Urbana-Champaign, USA.

Oct 13, 2026J/K move · Enter open · S save

WAVE-0: Bridging VLA Policies and WAMs with Lightweight Latent Future Prediction

Haomin Wang, Anqiong Ji, Ke Wang, Zhongshan Liu+3

AutoLab, School of Artificial Intelligence, Shanghai Jiao Tong University · Baidu AI Cloud · Awaken Robotics+1

World Models for Robotics World Action Models Vision-Language-Action Models Latent World Models

Open paper AskPDF arXiv

Vision-language-action (VLA) policies excel at generalist robotic manipulation but generate actions without explicitly modeling their environmental consequences. World action models (WAMs) incorporate future prediction, yet most rely on costly video generation, while latent alternatives either use predicted futures only as auxiliary targets or infer latent actions from suboptimal visual transitions through costly large-scale video pretraining. We introduce WAVE-0 (World-model-Augmented VLA Execution), a framework that bridges VLAs and WAMs through a lightweight latent world model. By connecting the semantic understanding of pretrained VLAs with explicit future prediction, WAVE-0 allows anticipated action consequences to directly guide action generation. Built on a pretrained VLA, its 257M world model translates VLM-generated latent action plans into predicted future representations that condition the action expert, unifying latent planning, future prediction, and action generation. The world model is trained directly on ground-truth robot actions with a counterfactual ranking objective, requiring neither pixel reconstruction nor large-scale video pretraining. WAVE-0 achieves success rates of 91.47% on RoboTwin 2.0 and 98.7% on LIBERO, along with 88.0%, 61.4%, and 75.0% on LIBERO-Plus, LIBERO-Pro, and LIBERO-Para, respectively. On real-world tasks, it improves the success rate of

π_{0.5}

from 38.3% to 73.3%. These results demonstrate strong manipulation performance and generalization across diverse distribution shifts.

Abstract

Explore similar work

WAVE-0: Bridging VLA Policies and WAMs with Lightweight Latent Future Prediction

LaWAM: Latent World Action Models for Efficient Dynamics-Aware Robot Policies

Think Like a World Model, Act Like a VLA: Distilling World-Model Representations into Compact Robot Policies