Faulty reward functions in the wild

9 years ago 14
Reinforcement learning algorithms can break in surprising, counterintuitive ways. In this post we’ll explore one failure mode, which is where you misspecify your reward function.
Read Entire Article