Inside self-improving RL: asset or burden?
Researchers audited a self-discovered RL algorithm to find when its learning history helps or hurts adaptation.
Researchers audited a self-discovered RL algorithm to find when its learning history helps or hurts adaptation.