---

This is a library of agents, environments and complete benchmarks that are used by the reinforcement learning (RL) community for empirical analysis. 

NOTICE: The RL-Library project is currently in a state of change. We are currently working on a new format and distribution system. Please consider these pages as a repository of RL-Glue 1.* and 2.0 codes. These codes are offered "as-is" and DO NOT support RL-Glue 3.0 and higher. Keep checking back for more news on the revamped version of RL-Library (The new RL-Library, which contains RL-Glue 3.0 compatible code, can be found here).

• Best Sellers Shelf [Under Construction]
  
• Add to the RL-Library
• Agent shelf
  
• Environment shelf
  
• Experiment shelf
• Project shelf
• RL-Library discussion page
  
• RL-Glue

The UofA Reinforcement Learning Library is a dynamic entity that requires the continual contribution of the RL community to grow  and develop. The majority of the entries were collected form the 2005 NIPS workshop on reinforcement learning benchmarks and bakeoffs. Many of the agents and environments were created by researches outside the UofA. If you have an agent, environment, or benchmark that should be included, please feel free to add it to the RL-Library!

If you have questions, comments, or concerns please email:

---

Special thanks to Vasant Honavar <http://www.cs.iastate.edu/~cs573x/> for help with the library banner.