- Rating
- 3
- Confidence
- 2
- Informativeness
- 3
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
The paper introduce AVM-Eval to evaluate action following and find that IDM performance differs between real and generated videos. They further construct AVM-Pref and propose ACWM-RM, showing improvements in visual quality and action following over the evaluated reward baselines in downstream RL experiments.
Critical strengths
- The paper proposes AVM-Eval to assess the reliability of IDM-based action following evaluation on real and generated videos, providing a benchmark for analyzing existing IDM rewards.
- The paper constructs AVM-Pref with action-mismatch and video-mismatch preferences to supervise action following and visual quality, respectively. Ablations show their complementarity and improved ranking performance through joint training.
- ACWM-RM combines dynamics pretraining with preference learning to evaluate visual quality and action following without reference videos. RL experiments on DreamDojo-2B show improvements over compared reward baselines in both aspects.
Critical weaknesses
- When there is a conflict between AF and VQ, which type of video does ACWM-RM tend to prefer? I'd like to know whether this preference will change as the quality gap widens.
- The authors used PSNR/LPIPS to construct video preference pairs during training, and the evaluation in Table 5 also includes these metrics. Although SSIM and FVD are also reported, visual quality assessments independent of these training metrics still need to be strengthened. Could the authors provide additional evaluations that do not rely on these metrics?
- Could the authors report the time required for ACWM-RM to compute rewards for each video clip and compare it to the time taken for reward computation by IDM-, metric-, and VLM-based methods? This would help readers assess the actual computational overhead of ACWM-RM in fine-tuning.
Questions for authors
See weaknesses.
6 comments
Response to reviewer CDW1 (2/3)
W2. Visual quality evaluation needs stronger evidence from metrics independent of the PSNR and LPIPS used to construct training preference pairs.
Thank you for this suggestion. We have added two metrics that are not used to construct our training preference pairs: FID and L2$\text{Latent}\text{Latent}$, outperforming the strongest baseline, metric-based, on both metrics. This indicates that the gains of ACWM-RM are not limited to the metrics used for preference supervision.
Both metrics are used to evaluate world models in IRASim [1] and OSCAR [2]. They assess visual quality from perspectives complementary to PSNR and LPIPS: FID measures distributional similarity between generated and real frames in Inception feature space, while L2$\text{Latent}$ measures per-sample fidelity in a VAE latent space. Following IRASim, we compute L2$\text{Latent}$ using the pretrained SDXL VAE as the latent encoder.
Table 3. Evaluation on metrics not used for preference supervision. L2$_\text{Latent}$ is reported ×100.
| Method | FID ↓ | L2$_\text{Latent}$ ↓ |
|---|---|---|
| Original | 5.01 ± 0.05 | 25.12 ± 0.03 |
| IDM-based | 5.39 ± 0.09 | 25.87 ± 0.14 |
| VLM-based | 5.47 ± 0.11 | 25.26 ± 0.07 |
| Metric-based | 4.30 ± 0.04 | 25.04 ± 0.06 |
| Ours | 4.25 ± 0.07 | 24.90 ± 0.06 |
[1] IRASim: A Fine-Grained World Model for Robot Manipulation. ICCV 2025.
[2] OSCAR: Omni-Embodiment Skeleton-Conditioned World Action Model for Robotics. arXiv 2026.
Related revisions: We have added FID and L2$_\text{Latent}$ results to Table 5 and updated the evaluation protocol in Section 5.1 (lines 364–367) and the visual-quality discussion in Section 5.3 (lines 420–427). All revisions are highlighted in red in the manuscript.
Response to reviewer CDW1 (3/3)
W3. How does ACWM-RM’s per-clip reward computation time compare with IDM-, metric-, and VLM-based methods, and what computational overhead does it add during fine-tuning?
We thank the reviewer for this valuable suggestion. In our experiments, ACWM-RM adds little overhead: it takes 0.54s to score a clip, which is only 7.42% of the total per-clip time. Its reward scoring is about 10× faster than that of the VLM-based method and 20× faster than that of the metric-based method, and only 0.44 s slower than that of the lightweight IDM-based method.
To measure the cost of reward scoring during fine-tuning, we record the per-clip time for rollout generation, reward scoring, and their sum for ACWM-RM and the IDM-, VLM-, and metric-based methods. We run all measurements on 100 DreamDojo-2B rollout clips (13 frames each), using the same dedicated H800 GPU with batch size 1.
We warm up the DreamDojo and each reward signal separately, then process the clips one at a time. Generation time is the wall-clock time DreamDojo-2B takes to produce one rollout clip. Reward-scoring time runs from when the generated clip is passed to the rewards until the score is returned. Reward share is reward-scoring time divided by total time.
Table 4. Per-clip generation and reward-scoring latency.
| Method | Generation (s) | Reward-scoring (s) ↓ | Total (s) ↓ | Reward share ↓ |
|---|---|---|---|---|
| IDM-based | 6.72 | 0.10 | 6.83 | 1.52% |
| VLM-based | 6.72 | 5.36 | 12.09 | 44.38% |
| Metric-based | 6.72 | 10.87 | 17.60 | 61.80% |
| Ours | 6.72 | 0.54 | 7.26 | 7.42% |
As shown in Table 4, generation time is identical across methods, so the differences come entirely from reward scoring. Reward scoring takes 44.38% and 61.80% of the per-clip time for the VLM-based and metric-based methods, making their total time 1.7× and 2.4× that of ACWM-RM. The IDM-based method is faster owing to its lightweight ResNet backbone, but it saves only about 6% of the total time and yields notably inferior performance.
Related revisions: We have added these experimental results to Appendix J (lines 1981–1991; Table 22). All revisions are highlighted in red in the manuscript.
Response to reviewer CDW1 (1/3)
Dear Reviewer CDW1,
We appreciate your time and constructive feedback, which has been invaluable in improving our work. We also express our sincere gratitude for your positive assessment of our paper and for recognizing the value of our contributions.
Thank you for this insightful question. Our new experiment shows that when AF and VQ conflict, ACWM-RM tends to prefer the higher-VQ video. This preference strengthens as the VQ gap widens and generally weakens as the AF gap widens.
Specifically, we build the test set from the pairs excluded when constructing the video-mismatch pairs of AVM-Eval, which require the positive sample to have a higher AF score than the negative sample and similar VQ scores. Among them, we keep the conflicting pairs, in which one video has a higher AF score, and the other has a higher VQ score. Following Appendix I, the AF gap is the nDTW difference (Eq. (51), line 1868), and the VQ gap is the root mean square of the differences in the four VQ metrics, each normalized by its corresponding standard deviation (Eq. (52), line 1800). We split each gap into small, medium, and large levels by its tertiles (Table 1), and randomly sample 250 pairs from each of the resulting nine cells, yielding 2,250 pairs.
Table 1. Gap thresholds.
Table 2. ACWM-RM’s preference rate (%) for the higher-VQ video.
As shown in Table 2, preference rates of ACWM-RM for the higher-VQ video range from 69.6% to 91.6% across all cells. This preference strengthens as the VQ gap widens (e.g., 69.6% → 88.8% under a large AF gap) and generally weakens as the AF gap widens (e.g., 82.8% → 69.6% under a small VQ gap), suggesting that ACWM-RM remains sensitive to AF.
We attribute this preference mainly to our training data. Video-mismatch pairs, which provide explicit VQ supervision, outnumber action-mismatch pairs, which provide explicit AF supervision. As a result, ACWM-RM may rely more on VQ cues when AF and VQ conflict.
Related revisions: We have added this experiment to Appendix J (lines 2018–2037; Tables 24 and 25). All revisions are highlighted in red in the manuscript.