Generative world models used as closed-loop test oracles require a five-level admissibility ladder (L0-L4) because visual fidelity does not predict action-robustness.
Do World Action Models Generalize Better than VLAs? A Robustness Study
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
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 $\pi_{0.5}$ can achieve comparable robustness on certain tasks, they typically require extensive training with diverse robotic datasets and varied learning objectives. Hybrid approaches that partially incorporate video-based dynamic learning exhibit intermediate robustness, highlighting the importance of how video priors are integrated.
citation-role summary
background 1
baseline 1
citation-polarity summary
years
2026 8representative citing papers
OA-WAM uses persistent address vectors and dynamic content vectors in object slots to enable addressable world-action prediction, improving robustness on manipulation benchmarks under scene changes.
Attacks can corrupt the latent future trajectory imagined by world-action models in VLA policies, causing failures in oracles like MPC while the reactive policy stays intact.
ImageWAM shows image editing models can replace video generation in world action models, delivering better performance with 6x lower FLOPs and 4x lower latency by using edit-derived KV caches as compact context.
Flash-WAM introduces modality-specific consistency parametrizations to distill joint video-action diffusion models to single-step inference, delivering 23x speedup with preserved benchmark performance.
X-WAM unifies robotic action execution and 4D world synthesis by adapting video diffusion priors with a lightweight depth branch and asynchronous noise sampling, achieving 79-91% success on robot benchmarks.
Survey organizing world models for robotic manipulation into representation families, a functional taxonomy, and infrastructure roles across pretraining, post-training, and inference, while reviewing 34 datasets and evaluation protocols.
MemoryWAM is a world action model with a hybrid memory design using recent frames, anchor frames, and gist tokens for efficient long-horizon robotic manipulation.
citing papers explorer
-
Validate the Dream Before You Trust Its Verdict: Admissibility for World-Model Simulators
Generative world models used as closed-loop test oracles require a five-level admissibility ladder (L0-L4) because visual fidelity does not predict action-robustness.
-
OA-WAM: Object-Addressable World Action Model for Robust Robot Manipulation
OA-WAM uses persistent address vectors and dynamic content vectors in object slots to enable addressable world-action prediction, improving robustness on manipulation benchmarks under scene changes.
-
Attacking the Trusted Imagination: Oracle-Level Integrity Attacks on Imagine-then-Act World Models
Attacks can corrupt the latent future trajectory imagined by world-action models in VLA policies, causing failures in oracles like MPC while the reactive policy stays intact.
-
ImageWAM: Do World Action Models Really Need Video Generation, or Just Image Editing?
ImageWAM shows image editing models can replace video generation in world action models, delivering better performance with 6x lower FLOPs and 4x lower latency by using edit-derived KV caches as compact context.
-
Flash-WAM: Modality-Aware Distillation for World Action Models
Flash-WAM introduces modality-specific consistency parametrizations to distill joint video-action diffusion models to single-step inference, delivering 23x speedup with preserved benchmark performance.
-
Unified 4D World Action Modeling from Video Priors with Asynchronous Denoising
X-WAM unifies robotic action execution and 4D world synthesis by adapting video diffusion priors with a lightweight depth branch and asynchronous noise sampling, achieving 79-91% success on robot benchmarks.
-
World Models for Robotic Manipulation: A Survey
Survey organizing world models for robotic manipulation into representation families, a functional taxonomy, and infrastructure roles across pretraining, post-training, and inference, while reviewing 34 datasets and evaluation protocols.
-
MemoryWAM: Efficient World Action Modeling with Persistent Memory
MemoryWAM is a world action model with a hybrid memory design using recent frames, anchor frames, and gist tokens for efficient long-horizon robotic manipulation.