References

CARL is built on top of other environments and benchmarks, extending them to contextual reinforcement learning environments.

Specifically, we used the works of:

OpenAI gym, Brockman et al., 2016. arXiv preprint arXiv:1606.01540

Brax – A Differentiable Physics Engine for Large Scale Rigid Body Simulation, Freeman et al., NeurIPS 2021 (Dataset & Benchmarking Track)

TOAD-GAN: Coherent Style Level Generation from a Single Example, Awiszus et al., AIIDE 2020

Learning to Design RNA, Runge et al., ICRL 2019

For insights into cRL, we recommend our CARL paper as well as:

Contextual Markov Decision Processes, Hallak et al., CoRR 2015

A survey of Generalisation in Deep Reinforcement Learning, Kirk et al., CoRR 2021

previous

License