Procgen Benchmark
OpenAI Blog
OpenAI released Procgen Benchmark, a set of 16 procedurally-generated environments designed to measure how quickly reinforcement learning agents learn generalizable skills. The benchmark includes 16 distinct environments that generate new levels procedurally to test agent generalization across novel scenarios. Researchers can now use this standardized tool to compare RL agent performance on learning efficiency and transferability rather than memorization of fixed levels.
Why it matters
We’re releasing Procgen Benchmark, 16 simple-to-use procedurally-generated environments which provide a direct measure of how quickly a reinforcement learning agent learns generalizable skills.
