Unpacking SDXL Turbo: Interpreting Text-to-Image Models with Sparse Autoencoders Paper β’ 2410.22366 β’ Published Oct 28, 2024 β’ 84