3 ms·
I was curious if anyone had tried to train an AI to play QWOP. Of course they have: https://www.youtube.com/watch?v=e27TUmMkOA0 https://www.youtube.com/watch?v=
by tlrobinson 8y ago
I was curious if anyone had tried to train an AI to play QWOP. Of course they have: https://www.youtube.com/watch?v=e27TUmMkOA0 https://www.youtube.com/watch?v=e27TUmMkOA0 (among others)
I wonder if you could get a better result by including other factors in the reward function, like trying to maintain a slight forward lean.