Center for AI Safety — publication: Representation Engineering: A Top-Down Approach to AI Transparency — proposes methods to read and control LLM internal representations for safety

Note: The source directly confirms all key elements of the claim: (1) CAIS is listed as the affiliation for multiple authors including the first author Andy Zou; (2) the publication title matches exactly; (3) the abstract describes methods for 'monitoring and manipulating' (i.e., reading and controlling) LLM internal representations; (4) safety applications are explicitly mentioned including 'harmlessness' and 'power-seeking'; (5) the arxiv ID 2310.01405 corresponds to October 2023, matching the 'as of 2023-10' temporal specification. All author names listed in the claim appear in the source document.