- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
MoS-VLA learns 16 image-and-language-conditioned basis functions on an OpenVLA backbone. It adapts to a new context by fitting their coefficients to demonstration actions through convex L1 regression, with no model weight updates. The paper evaluates action prediction, adaptation cost, and robot control. Appendix H extends the method to latent action chunks.
Critical strengths
The adaptation step is simple and practical. Table 2 shows much lower adaptation time and GPU memory use than gradient-based finetuning, with competitive prediction error at 128 examples.
The paper includes simulation and real-robot results. The calibrated single-head control in Figure 4 is also useful: the multiple-basis model lowers offline error on eight of ten held-out datasets.
Critical weaknesses
-
The main robot comparison lacks matched adaptation controls. Table 1 includes zero-shot OpenVLA and pi0.5, plus ICRT and X-VLA, but no one-demonstration OFT/LoRA baseline or calibrated single-head control. These controls appear only in offline evaluations. The results therefore leave open whether the proposed adaptation method offers a better tradeoff between adaptation cost and closed-loop success.
-
One-shot reliability is not tested across independent calibration demonstrations. Repeated rollouts of one calibrated policy do not show how sensitive adaptation is to the chosen demonstration. Tables 2 and 6 vary sample counts, not trajectory selection or state coverage. The limited task variation and disabled forward/backward motion in three tasks further narrow the evidence for generalization.
-
Figure 6 shows uneven approximation quality across unseen domains, but does not link projection residuals to control performance. It remains unclear whether a high residual identifies contexts where coefficient fitting will fail, or when additional demonstrations can help. This matters because adaptation cannot change the learned function span.
-
The latent extension provides limited evidence for the added basis. At 128 examples, calibration alone gives L1 error 0.369, while the 0.184 result also finetunes the VAE. The authors clearly distinguish these settings. However, the matched VAE without basis functions reaches 0.189 (Table 7), leaving only a small average gain without uncertainty estimates. The stronger latent result therefore does not establish a gradient-free improvement.
Questions for authors
- Can the calibrated single-head model and one-demonstration OFT/LoRA be included in the robot comparison?
- How were calibration trajectories selected, and which initial states or scene conditions change during testing?
- Are fitted coefficients shared across language tasks in the same lab? Are finetuned VAE weights shared across target contexts?
Thank you for the detailed review and for highlighting our offline efficiency and k=1 gains. We have now addressed the closed-loop concern with LIBERO-Plus (200 trials/cell, Appendix G) and Franka block-lifting baselines, where MoS-VLA reaches 10/10 in under 10s while calibrated k=1 and 1-demo OpenVLA-OFT both score 0/10.
Q1. Calibrated single-head (k=1) and 1-demo OFT/LoRA in robot comparisons
MoS-VLA outperforms k=1 in LIBERO-Plus and beats both k=1 (0/10) and 1-demo OpenVLA-OFT (0/10) on real-robot block lifting:
LIBERO-Plus (Appendix G, Table 5): Across 9 contexts (200 trials/cell), MoS-VLA beats calibrated k=1 in-distribution (0.47 vs. 0.32) and on held-out compositions (0.37-0.53 vs. 0.14-0.27).
Real-robot block lifting: MoS-VLA achieves 10/10 in under 10s; calibrated k=1 and 1-demo OpenVLA-OFT (500 steps, about 15 min) both get 0/10, while 1-demo LoRA finetuning of π0.5 reaches 10/10 only after 500 gradient steps (about 15 min).
Q2. Calibration trajectory selection and test-time initial-state variations
Demo-selection sensitivity: Calibration uses one uncurated demonstration per task. Varying which training episode is used within each of the 10 OOD datasets yields an average within-dataset std of only 0.027 around the 0.265 mean L1 (0.245±0.017 with ∣α∣≤1). Using the first episode (as in the main paper, 0.274 L1) is slightly worse than the average across episodes (0.265), confirming we did not cherry-pick a favorable demo.
Initial-state and scene variations: In LIBERO-Plus, calibration and evaluation episodes are sampled independently from LIBERO's initial-state distribution across 200 trials/cell and 9 scene/viewpoint shifts.
Q3 & W3-W4. Coefficient/VAE sharing, projection residuals (Fig. 6), and latent basis (App. I)
Coefficient sharing:
VAE weights (Appendix I):
Projection residuals (Fig. 6): Figure 6 measures whether a target task can be expressed by combining the 16 learned basis functions at all, and LIBERO-Plus confirms how this predicts control performance: