2 ms·How will you train your world model? Cross-entropy loss? Oh, it's reward again.by canjobear 5y agoHow will you train your world model? Cross-entropy loss? Oh, it's reward again.