Posts tagged #Ai

A paper on pruning transformer attention heads during inference without retraining.