Upper Confidence Bound

Concept

A standard reinforcement learning algorithm used by ERA to optimistically select which code mutations to explore, aiming for efficient progress rather than immediate best performance.

Mentioned in 1 video

Videos Mentioning Upper Confidence Bound

🔬 Google's AI Scientist Started as an Attempt to Automate Kaggle — John Platt, Google Fellow

🔬 Google's AI Scientist Started as an Attempt to Automate Kaggle — John Platt, Google Fellow

Latent Space

A standard reinforcement learning algorithm used by ERA to optimistically select which code mutations to explore, aiming for efficient progress rather than immediate best performance.