Quantifying generalization in reinforcement learning

OpenAI · 7 years ago · 29

Researchers released CoinRun, a training environment designed to measure how well reinforcement learning agents transfer knowledge to unfamiliar scenarios. The environment sits between simple benchmarks and complex games like Sonic the Hedgehog, offering a calibrated test for generalization capabilities. This tool has resolved previous debates about why some reinforcement learning algorithms struggle with transfer learning despite strong performance on training tasks.