no code implementations • 27 May 2023 • Omkar Ranadive, Nikhil Thakurdesai, Ari S Morcos, Matthew Leavitt, Stéphane Deny
Finally, in causal experiments where we regularize against class selectivity at different points in training, we show that the presence of class-selective neurons early in training is critical to the successful training of the network; in contrast, class-selective neurons can be suppressed later in training with little effect on final accuracy.