Dreamer-MCX

Official implementation of "Exploiting Model Errors for Exploration in Model-Based Reinforcement Learning", accepted to EWRL18.

Acknowledgments

This code is built directly on top of the this implemenation of DreamerV3 in Pytorch.

Citation

@misc{swift2025exploiting,
    title = Exploiting Model Errors for Exploration in Model-Based Reinforcement Learning
    author = {Swift, Jared and Leonetti, Matteo},
    howpublished = {Presented at the 18th European Workshop on Reinforcement Learning (EWRL)},
    year = {2025}
}