Add to Favourites
To login click here
A team of researchers trained a neural network using reinforcement learning to play the popular roguelike game NetHack. However, the model suddenly performed worse on one run, prompting the researchers to investigate the cause.
