Comment by khafra

LLM model interpretability also uses Sparse Autoencoders to find concept representations (https://openai.com/index/extracting-concepts-from-gpt-4/), and, more recently, linear probes.