ReLAx - Reinforcement Learning Applications Library

Study Model-Based Policy Optimization by varying the model estimator classes (e.g Decision Trees vs MLP)

Numerical Evidence for Sample Efficiency of Model-Based over Model-Free Reinforcement Learning Control of Partial Differential Equations [ECC'24]

Example MBPO implementation with ReLAx

Model-Based Policy Optimization in JAX/FLAX/SBX

A re-implementation of MBPO algorithm for MDP env.