Interpreting and Steering Protein Language Models through Sparse Autoencoders Paper • 2502.09135 • Published Feb 13 • 1