- Rating
- 3
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
This paper studies the vulnerability of interactive visual world models, where even a slight perturbation to the initial input can lead to eventual failure. It proposes a label-free adversarial method to achieve consistent attacks under diverse future controls. Importantly, experiments in autonomous driving scenarios show that such attacks can lead to severe safety consequences.
Critical strengths
- This paper is well written and easy to follow.
- The studied problem is important, i.e., robustness in interactive visual world models.
Critical weaknesses
- The proposed method assumes that the attacker has full model access, which is unrealistic in real-world deployment scenarios. When the model is inaccessible, it is suggested that the authors discuss the possible transferability of the method to strengthen its value.
- The authors claim that a major challenge for the attack is obtaining the future ground-truth video. Considering that the ultimate goal of the attack is to corrupt the generated results, why cannot Equation 4 directly align Gθ(x + δ, τ1:K) with random noise or with the results under the opposite control? Therefore, the ground-truth video can be easily bypassed rather than being a key challenge.
- The authors claim that the perturbations are imperceptible. However, the evaluation only considers human-level perception, and it remains unclear whether the perturbations introduce detectable statistical artifacts. A simple adversarial detector or perceptual analysis could help quantify whether the perturbations are genuinely imperceptible rather than merely visually subtle.
Questions for authors
Please refer to the issues raised in the Weaknesses section.
Response to Reviewer DSmP
We thank the reviewer for recognizing the clarity of our presentation and the importance of robustness in visual world models. We will address your concerns in the following response.
>>> W1Transferability Without Target-Model Access>>> A1We appreciate this practical suggestion. We completely agree that transferability is important when the target model is inaccessible. To prove this, we present the existing black-box transfer results fromAppendix A.3and add another cross-model evaluation.Appendix A.3evaluates transfer between Matrix-Game 2.0 and Astra. Adversarial images are resized to the target model’s input resolution before inference.The Matrix-Game → Astra transfer reduces Background Consistency from 0.878 to 0.846 and Imaging Quality from 0.553 to 0.515 without re-optimizing on Astra. This confirms that BadWorld perturbations remain effective on an unseen target model.
The transferred attacks degrade all four metrics, reducing Imaging Quality from 0.697 to 0.609 and CLIP-I from 0.885 to 0.803. The additional evaluation further supports transferability beyond the white-box model used for optimization.
>>> W2Targeted Objectives Still Need a Rollout Proxy>>> A2Thanks for raising this thoughtful question! We believe there is a misunderstanding here. A random-noise or opposite-control video can provide target supervision for a targeted attack, but it cannot replace the video latent required as denoising input.Specifically, under a training-style diffusion denoiser, we must first construct a noisy latent by adding noise to a clean video latent. However, in our setting, only the context image is available; this ground-truth video latent is missed. Changing the supervision target therefore changes only the loss and still leaves us without a valid denoising input or autoregressive history condition. BadWorld directly solves this by using repeated context frames as a surrogate history and self-predicted velocity as label-free supervision.
>>> W3Imperceptibility>>> A3Thanks! We agree that visual subtlety alone does not rule out detectable statistical artifacts. To first quantify the perturbation magnitude, as stated inline 413, our default perturbation budget is 0.05 in the normalized pixel range [−1,1]. This corresponds to 0.025 in [0,1], or approximately 6.375/255.(Arrows indicate better preservation of the original image.)
These results show that BadWorld introduces only small perceptual changes and maintains high fidelity to the clean input.