Revisiting End-To-End Sparse Autoencoder Training: A Short Finetune Is All You Need

Re-assign community