2 ms·
Reinforcement learning was applied after the basic model was initialized with imitation. Maybe that can partially explain the small number of steps.
by caxap 16y ago
Reinforcement learning was applied after the basic model was initialized with imitation. Maybe that can partially explain the small number of steps.