Sparsity
Neural networks don't want to be black boxes, they just need reminding of this. An approach to ML training that increases model speed, interpretability, and control. Tested across 8 OOM of scale.
Neural networks don't want to be black boxes, they just need reminding of this. An approach to ML training that increases model speed, interpretability, and control. Tested across 8 OOM of scale.
Log in or sign up for Devpost to join the conversation.