RL-Library is a collection of agents, environments, and experiments written in a variety of languages that are compatible with RL-Glue. The selection is currently a bit limited (Mountain Car, Tetris, and a few others), but over time we hope that RL-Library can provide one-stop shopping for finding vetted test-problems for new Reinforcement Learning Algorithms.
The semi/sortof annual Reinforcement Learning Competition.
A pet project created and maintained by Brian Tanner. Includes facilities to do direct-compile experiments with a mix of Java and C/C++ agents/environments/experiments. Also has facilities for dynamic, run-time loading of agents and environments, configurable/parametrized agents/environments, and agent/environment visualization. RL-Viz was the base for the 2008 RL-Competition software.
Your Project Here!
If you have a project related to RL-Glue, tell us about it! We can offer you space on this site to talk about it!