MyoSuite

MyoSuite is a benchmark dataset that contains a variety of continuous-control tasks used to evaluate the performance of model-based reinforcement learning algorithms.

Papers using MyoSuite (5)

WIMLE: Uncertainty-Aware World Models with IMLE for Sample-Efficient Continuous Control2026Scaling Off-Policy Reinforcement Learning with Batch and Weight Normalization2025SimBa: Simplicity Bias for Scaling Up Parameters in Deep Reinforcement Learning2024Reward-free World Models for Online Imitation Learning2024Bigger, Regularized, Optimistic: scaling for compute and sample-efficient continuous control2024 · 1 cites