- Rating
- 1
- Confidence
- 4
- Informativeness
- 1
- Presentation
- 2
- Soundness
- 1
Read review
Paper summary
UATL keeps the original VLA backbone, action-chunking protocol, query frequency, and open-loop execution. It adds (i) Heteroscedastic Action Learning (HAL), which predicts a Laplace location and a per-time-step/per-dimension scale, and (ii) an uncertainty-weighted second-order temporal-consistency regularizer (UW-TCR). The inverse predicted scales weight the smoothness penalty, so confident regions are smoothed more strongly while uncertain regions can change rapidly. Experiments use OpenVLA-7B with LoRA. UATL reports 98.3% average success on LIBERO, 71.2% on LIBERO-Plus, 69.3%/54.3% on the two Google Robot settings of SimplerEnv, 63.5% on WidowX, and an average CALVIN sequence length of 4.05 (96.0/89.6/81.0/73.8/65.0% success for lengths 1--5). On LIBERO, the ablation improves 95.8% (baseline) to 96.6% (HAL) and 98.3% (HAL+UW-TCR), with only a small inference overhead.
Critical strengths
- The method is simple to integrate: it changes the action head and training loss without changing the backbone, policy-query schedule, or execution pipeline.
- HAL and UW-TCR are conceptually complementary and address both nonuniform prediction reliability and within-chunk temporal jitter.
- The evaluation is broad, covering standard manipulation, perturbation robustness, cross-platform transfer, and long-horizon control. The ablation and throughput measurements help isolate the proposed components.
Critical weaknesses
- The paper treats the predicted Laplace scales as aleatoric uncertainty, but does not report calibration metrics, reliability diagrams, scale-error correlations, or comparisons with calibrated uncertainty baselines. It is therefore unclear whether the scales are meaningful uncertainty estimates or simply adaptive loss weights.
- The methodological novelty is incremental relative to heteroscedastic regression, jerk/acceleration regularization, and action-chunk continuity or aggregation methods. Strong matched comparisons with RTC, overlapping-chunk aggregation, and recent temporal-smoothing methods are missing.
- LIBERO is close to saturation, and the absolute gain over the baseline is 2.5 points. Results are less consistent under harder transfer conditions, especially Variant Aggregation (54.3%) and drawer tasks. The paper should avoid attributing all gains to uncertainty modeling without stronger transfer evidence.
- There is no real-robot evaluation. A second-order penalty could suppress necessary abrupt contact transitions or gripper-adjacent changes, even if the current simulation tasks do not expose this failure mode.
- Results rely on OpenVLA-7B, LoRA, and a selected regularization weight (lambda_TCR=0.05). More seeds, confidence intervals, and robustness across backbones and training choices are needed to establish that the gains are not setup-specific.
Questions for authors
- Please report NLL, calibration error/reliability plots, rank correlation between predicted scales and realized action errors, and results with shuffled or detached random scales. This would test whether HAL learns uncertainty rather than only a reweighting heuristic.
- Please compare against strong temporal baselines (first/second-order smoothness, RTC or action aggregation) using the same OpenVLA checkpoint, data, training steps, and hyperparameter-selection budget.