The greatest ambition of mechanistic interpretability is to completely rewrite deep neural
networks in a format that is more amenable to human understanding, while preserving their …

Automated interpretability pipelines generate natural language descriptions for the concepts
represented by features in large language models (LLMs), such as plants or the first word in …

Sparse autoencoders (SAEs) are a useful tool for uncovering human-interpretable features
in the activations of large language models (LLMs). While some expect SAEs to find the true …

Mechanistic interpretability is the program of explaining what AI systems are doing in terms
of their internal mechanisms. I analyze some aspects of the program, along with setting out …

Sparse autoencoders (SAEs) are an increasingly popular technique for interpreting the
internal representations of transformers. In this paper, we apply SAEs to'interpret'random …

Effective and reliable control over large language model (LLM) behavior is a significant
challenge. While activation steering methods, which add steering vectors to a model's …