Does anyone know of a Julia package which wraps the Arcade Learning Environment (http://www.arcadelearningenvironment.org/)? I found @Andy-P's package https://github.com/Andy-P/DeepQLearning.jl, but this only has the learning algorithm, not the underlying framework. My interest lies in novel approaches to AGI, and a Julia-wrapped ALE would enable some good benchmarks. Also if anyone else shares these goals and is interested in collaboration, please reach out.
Thanks, Tom