RLCD

Is Reinforcement Learning Really Needed for Jev's Spreadsheet AI?

The question cuts to the heart of Jev's design.

4 min readMachine Learning
From Machine Learning

Just saw the YouTube presentation and I was left wondering this question. If jev only outputs Choice, Score, or Noul … well those are all perfectly differentiable. (Cross entropy or mse) I don’t know if I’m missing something or if adding RL is just for marketing. Like what would an RL environment even look like?

Read the original at Machine Learning