# Robot Data Curation with Mutual Information Estimators

arXiv: https://arxiv.org/abs/2502.08623 · HTML: https://mlsanity.com/paper/2502.08623/

## Key facts

- Cited by 37 papers in the map (data through 14 Oct 2026).
- Placed in Robot manipulation imitation learning, one of 13 directions in Offline RL and imitation learning from data.
- First posted on arXiv 12 Feb 2025.

## Placed in

- Offline RL and imitation learning from data / [Robot manipulation imitation learning](https://mlsanity.com/offlinerl/robot-manipulation-imitation-learning/)

## Built on by (most cited first, 30 of 37)

- [π0.7: a Steerable Generalist Robotic Foundation Model with Emergent Capabilities](https://mlsanity.com/paper/2604.15483/) (2026)
- [Robometer: Scaling General-Purpose Robotic Reward Models via Trajectory Comparisons](https://mlsanity.com/paper/2603.02115/) (2026)
- [Shortcut Learning in Generalist Robot Policies: The Role of Dataset Diversity and Fragmentation](https://mlsanity.com/paper/2508.06426/) (2025)
- [DataMIL: Selecting Data for Robot Imitation Learning with Datamodels](https://mlsanity.com/paper/2505.09603/) (2025)
- [CUPID: Curating Data your Robot Loves with Influence Functions](https://mlsanity.com/paper/2506.19121/) (2025)
- [World Action Verifier: Self-Improving World Models via Forward-Inverse Asymmetry](https://mlsanity.com/paper/2604.01985/) (2026)
- [Guiding Data Collection via Factored Scaling Curves](https://mlsanity.com/paper/2505.07728/) (2025)
- [GeoSem-WAM: Geometry- and Semantic-Aware World Action Models](https://mlsanity.com/paper/2606.03188/) (2026)
- [Data Retrieval with Importance Weights for Few-Shot Imitation Learning](https://mlsanity.com/paper/2509.01657/) (2025)
- [SCIZOR: A Self-Supervised Approach to Data Curation for Large-Scale Imitation Learning](https://mlsanity.com/paper/2505.22626/) (2025)
- [FrameSkip: Learning from Fewer but More Informative Frames in VLA Training](https://mlsanity.com/paper/2605.13757/) (2026)
- [Using Non-Expert Data to Robustify Imitation Learning via Offline Reinforcement Learning](https://mlsanity.com/paper/2510.19495/) (2025)
- [Ambient Diffusion Policy: Imitation Learning from Suboptimal Data in Robotics](https://mlsanity.com/paper/2606.12365/) (2026)
- [Uncertainty Quantification for Flow-Based Vision-Language-Action Models](https://mlsanity.com/paper/2606.18043/) (2026)
- [Counterfactual Behavior Cloning: Offline Imitation Learning from Imperfect Human Demonstrations](https://mlsanity.com/paper/2505.10760/) (2025)
- [RDD: Retrieval-Based Demonstration Decomposer for Planner Alignment in Long-Horizon Tasks](https://mlsanity.com/paper/2510.14968/) (2025)
- [Learning from the Best: Smoothness-Driven Metrics for Data Quality in Imitation Learning](https://mlsanity.com/paper/2604.23000/) (2026)
- [AttenA+: Rectifying Action Inequality in Robotic Foundation Models](https://mlsanity.com/paper/2605.13548/) (2026)
- [Tri-Info: Generalizable, Interpretable Failure Prediction for VLA Models via Information Theory](https://mlsanity.com/paper/2606.19998/) (2026)
- [An Efficient Metric for Data Quality Measurement in Imitation Learning](https://mlsanity.com/paper/2605.01544/) (2026)
- [FACTR 2: Learning External Force Sensing for Commodity Robot Arms Improves Policy Learning](https://mlsanity.com/paper/2606.12406/) (2026)
- [Z-1: Efficient Reinforcement Learning for Vision-Language-Action Models](https://mlsanity.com/paper/2606.31846/) (2026)
- [Retrieve-Augmented Generation for Speeding up Diffusion Policy without Additional Training](https://mlsanity.com/paper/2507.21452/) (2025)
- [Quality over Quantity: Demonstration Curation via Influence Functions for Data-Centric Robot Learning](https://mlsanity.com/paper/2603.09056/) (2026)
- [ATHENA: Accelerated Multi-Task Heterogeneous Influence Functions for Robot Data Curation](https://mlsanity.com/paper/2606.16208/) (2026)
- [Active Real-World Factor-Based Evaluation for Generalist Robot Policies](https://mlsanity.com/paper/2607.14439/) (2026)
- [Detecting Mislabeled and Corrupted Data via Pointwise Mutual Information](https://mlsanity.com/paper/2508.07713/) (2025)
- [Mechanisms of Misgeneralization in Physical Sequence Modeling](https://mlsanity.com/paper/2605.20299/) (2026)
- [TSD: A Physics-Inspired Trajectory Saliency Detector for Efficient Imitation Learning](https://mlsanity.com/paper/2606.23371/) (2026)
- [SIEVE: Structure-Aware Data Selection for Imitation Learning with VLA Models](https://mlsanity.com/paper/2607.06442/) (2026)

## Abstract (arXiv)

The performance of imitation learning policies often hinges on the datasets with which they are trained. Consequently, investment in data collection for robotics has grown across both industrial and academic labs. However, despite the marked increase in the quantity of demonstrations collected, little work has sought to assess the quality of said data despite mounting evidence of its importance in other areas such as vision and language. In this work, we take a critical step towards addressing the data quality in robotics. Given a dataset of demonstrations, we aim to estimate the relative quality of individual demonstrations in terms of both action diversity and predictability. To do so, we estimate the average contribution of a trajectory towards the mutual information between states and actions in the entire dataset, which captures both the entropy of the marginal action distribution and the state-conditioned action entropy. Though commonly used mutual information estimators require vast amounts of data often beyond the scale available in robotics, we introduce a novel technique based on k-nearest neighbor estimates of mutual information on top of simple VAE embeddings of states and actions. Empirically, we demonstrate that our approach is able to partition demonstration datasets by quality according to human expert scores across a diverse set of benchmarks spanning simulation and real world environments. Moreover, training policies based on data filtered by our method leads to a 5-10% improvement in RoboMimic and better performance on real ALOHA and Franka setups.