Interpreting and Steering Protein Language Models through Sparse Autoencoders

Open in new window