- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
This paper presents Enfold, a framework that transfers the intermediate computation of a video world model into a current-observation representation. During training, a predictive encoder learns to match multi-level generator states extracted from observed future videos, while the resulting representation also conditions future generation and a detached action head. At inference, control uses only the encoder and action head, bypassing video generation. Experiments on LIBERO, RoboTwin 2.0, and real-robot tasks show competitive control performance with substantially reduced action latency.
Critical strengths
-
Enfold introduces an interesting representation-level interface for world models, distilling multi-level generator states into a current-observation representation and avoiding expensive video generation during control.
-
The paper provides broad validation across simulation, real-robot control, future-video prediction, and representation analysis, showing competitive task performance with substantially lower inference latency.
Critical weaknesses
-
The control improvements are marginal. Enfold improves over Fast-WAM by only 0.2 points on LIBERO and performs slightly worse on RoboTwin 2.0, while Enfold-Flash differs by similarly small margins.
-
The reported efficiency gains are not fully isolated from model size and implementation differences. Enfold has roughly half the parameters of Fast-WAM, while the 10.1× Enfold-Flash result additionally uses TensorRT acceleration that is not applied to the baselines. A matched latency comparison using the same backbone, action head, sampling configuration, and deployment backend is needed to determine how much of the speedup comes specifically from bypassing the generative pathway.
Questions for authors
-
Can the authors provide results across multiple training seeds or confidence intervals for LIBERO and RoboTwin 2.0? The differences from Fast-WAM are only 0.2 points or less, so their statistical reliability would materially affect the assessment.
-
Can the authors provide a matched latency comparison that controls for model size, action-head configuration, sampling steps, and deployment backend? In particular, how much speedup is obtained solely by bypassing the video generator, separately from the smaller model and TensorRT acceleration?
2 comments
## Response to Reviewer kS4p (W2 / Q2)
W2 / Q2 — Inference architecture, action-head capacity, and deployment acceleration
Enfold removes the video generator from the deployed action path and makes the action head architecturally independent. Fast-WAM already skips future-video generation but still executes a video DiT to obtain action context. Its Mixture-of-Transformers (MoT) architecture couples the action expert to the video backbone through shared attention, constraining action-head depth and compatible attention interfaces. Enfold transfers predictive supervision into a current-observation encoder, so control requires only this encoder and the action head. This decoupling allows action-head depth, hidden dimension, and attention-head configuration to be chosen independently of the video backbone.
The submitted measurements on an A100 40GB separate the reported comparisons as follows:
| Comparison | Latency per action prediction | Observed ratio | Interpretation |
|---|---|---|---|
| Fast-WAM vs. standard Enfold | 493 ms vs. 134 ms | 3.68× | End-to-end comparison of different model configurations |
| Standard Enfold vs. Enfold-Flash | 134 ms vs. 49 ms | 2.73× | Deployment acceleration of the same Enfold model and inference path |
| Fast-WAM vs. Enfold-Flash | 493 ms vs. 49 ms | 10.06× | Combined architecture/configuration and deployment comparison |
The standard configuration delivers a 3.7× system-level speedup, and TensorRT provides a further 2.7× acceleration within Enfold, yielding the reported 10.1× overall ratio. The 3.7× comparison includes model-configuration differences; the 10.1× comparison additionally includes the deployment backend.
The Wan-teacher variant achieves 97.4% on LIBERO and the robustness results above, demonstrating that the representation interface extends beyond Cosmos. With the encoder and action head held fixed, changing the training-time teacher leaves the deployed computation graph unchanged.
Additional profiling clarifies the three sources of acceleration. We compared action-head depths and progressively enabled TensorRT. Ratios below use this new profiling run only.
| Configuration | Backend | Latency per action chunk |
|---|---|---|
| Fast-WAM | PyTorch | 530.26 ms |
| Enfold, 30-layer action head | PyTorch | 301.87 ms |
| Enfold, 10-layer action head (default) | PyTorch | 120.31 ms |
| Enfold, action head accelerated | PyTorch + TensorRT | 45.44 ms |
| Enfold, encoder and action head accelerated | TensorRT | 36.00 ms |
The 30-layer Enfold control is 1.76× faster than Fast-WAM in PyTorch, reflecting the overall architectural difference rather than isolating video-backbone removal alone. Beyond removing the video backbone, decoupling permits a smaller action head, and our predictive representation supports strong control at this reduced capacity: the 10-layer head adds 2.51× acceleration while achieving 97.8% LIBERO success versus Fast-WAM’s 97.6%. Finally, TensorRT adds 3.34× deployment acceleration (120.31 → 36.00 ms), separately from these architectural gains.
## Response to Reviewer kS4p (W1 / Q1)
We thank the reviewer for the questions on performance reliability and inference efficiency. Enfold's contribution is a representation interface that combines competitive control with improved robustness and low deployment cost. We have added independent training runs and distribution-shift evaluations, and clarify the sources of the reported speedup below. We have also included all the experiments in the revised version of the manuscript, with the added or revised content highlighted in red.
W1 / Q1 — Control performance across training seeds and distribution shifts
Enfold’s clearest advantage is under distribution shift, with substantial gains on LIBERO-Plus and improved RoboTwin clean-to-randomized transfer. Average LIBERO-Plus success improves from 51.5% for Fast-WAM to 68.7% with Cosmos (+17.2 percentage points) and 65.2% with Wan (+13.7 points). Enfold–Cosmos improves over Fast-WAM in six of seven LIBERO-Plus categories, including +43.3 points on Camera and +23.1 on Language. The Wan variant shows the same six-category improvement pattern, extending the robustness advantage across teachers.
These results reveal larger differences than those on saturated standard benchmarks.
For RoboTwin C2R, all methods are trained on clean demonstrations and tested on randomized scenes, with clean performance reported for reference. Our C2R training uses no data augmentation, which may contribute to the low absolute success under randomization:
Enfold–Cosmos and Fast-WAM perform similarly on clean scenes, but randomized success increases from 1.9% to 3.9% with Cosmos and 7.1% with Wan. Absolute success remains low in this challenging setting, but both OOD evaluations show an advantage beyond comparable in-distribution control.
To assess the reliability of standard-benchmark performance, we also independently retrained Enfold with seeds 42, 43, and 44. We report success (%) and mean ± sample standard deviation.
Enfold consistently maintains high success across independent training runs, with standard deviations of 0.06, 0.12, and 0.13 points. This supports the repeatability of its competitive standard-benchmark performance; its overall RoboTwin mean is 91.81%, comparable to Fast-WAM’s 91.83%.