Beyond Validation Loss: Coverage and Reliability Dynamics in Robot Policy Pretraining
Organizations: Stanford University · Carnegie Mellon University
Abstract
In modern robot learning, generative robot policies are first pretrained through supervised learning from demonstrations, then refined through reinforcement learning, supervised finetuning, or other forms of posttraining. Yet conventional pretraining metrics, such as validation loss and average task success, provide an incomplete account of which policies support both effective direct deployment and subsequent posttraining. In this work, we study closed-loop pass@k, which evaluates the probability that k rollouts from the same initial condition attain at least one success. Thus, pass@1 measures mean task success; large-k pass@k measures whether success remains reachable under repeated sampling. Across a wide range of posttraining recipes, we show that large-k pass@k reveals posttraining differences not captured by pass@1. In particular, checkpoints with similar pass@1 but different large-k pass@k have different posttraining performance. Moreover, we show pass@k peaks at different checkpoints during supervised pretraining, revealing a coverage-reliability tradeoff over the course of training. Based on these observations, we propose two practical interventions: (1) a purely offline checkpoint-selection criterion that balances coverage and reliability without closed-loop evaluation, and (2) a generative data augmentation procedure that expands supervision beyond the finite demonstration set, preserving large-k pass@k while maintaining or improving pass@1 and raising the observed coverage-reliability Pareto frontier. Together, our results motivate evaluation, checkpoint selection, and training procedures that explicitly account for posttraining. Blog post, visualizations, and code are available at https://zhanyisun.github.io/robot-policy-learning-dynamics/.
Figures & tables
| Task | Selected | Final |
|---|---|---|
| Threading (D2) | 0.371 / 0.886 | 0.321 / 0.812 |
| Mug Cleanup (D1) | 0.558 / 0.914 | 0.502 / 0.866 |
| Rollouts | PSNR | LPIPS |
|---|---|---|
| Chunk 1 | ||
| Selected | 27.9 | 0.112 |
| Final | 26.8 | 0.120 |
| Chunk 2 | ||
| Selected | 25.2 | 0.136 |
| Final | 24.1 | 0.145 |
Appendix figures & tables27 assets
Supplementary material from the paper’s appendix.
Appendix
| Training rollouts | |||
|---|---|---|---|
| Selected | 0.0800 | 0.1033 | 0.1186 |
| Final | 0.0815 | 0.1059 | 0.1219 |
| Hyperparameter | Value |
|---|---|
| Observation steps | 1 |
| RGB inputs | 2 views, (6-channel stack) |
| Proprioceptive state | 9-d |
| Predicted action chunk size | 8 |
| Executed action chunk size | 8 |
| Visual encoder | |
| Hyperparameter | Value |
|---|---|
| Image | 2 views, (6-channel stack) |
| Proprioceptive state | 9-d |
| Action chunk | 8 steps, absolute pose, 6-d rotation (10-d) |
| Observation steps | 1 |
| Image stream | |
| Architecture | conv U-Net |
| Mug | Square | Threading | Hammer | 3-Piece | 3-Piece | |
| D1 | D2 | D2 | D1 | D1 | D2 | |
| Demonstrations | 200 | 200 | 200 | 200 | 200 | 500 |
| Generator checkpoints | 100, 200, | 100, 200, | 100, 200, | 100, 200, | 100, 200, | 50, 150, |
| (training epoch) | 300, 400, | 300, 400 | 300, 400, | 300, 400, | 300, 400, | 250, 350, |
| 500 | 500 | 500 | 500 | 500 | 450 |
| Hyperparameter | Value |
|---|---|
| Ensemble members | |
| Ensemble size | 20 |
| Member architecture | policy observation encoder + MLP head |
| Head width / layers | 512 / 3 |
| Member output | flattened action chunk ( -d) |
| Ensemble optimization | |
| Hyperparameter | Value |
|---|---|
| Frozen base policy | EMA weights, DDIM 10 steps ( ), chunk |
| Residual actor | MLP , GELU, LayerNorm |
| Critic ensemble | MLP , GELU, LayerNorm |
| BC weight | 200 |
| TD loss / -step | MSE / chunks |
| Discount / target update | 0.99 / 0.01 |
| Hyperparameter | Value |
|---|---|
| Peak learning rate | / |
| Schedule | 1k-step warm-up, cosine decay to peak |
| Training steps | 100k / 50k |
| Global batch size | 256 |
| Optimizer | AdamW ( , , wd ), grad clip 1.0 |
| EMA decay (evaluated weights) | 0.99 |
Robot Learning
Robot learning develops methods for robots to acquire skills and adapt from demonstrations, interaction, data, and physical models.
See all 340 papers