1 paper
Marcelo Maciel Amaral, Raymond Aschheim
Large language models (LLMs) remain broadly open and highly steerable: they imitate at scale, accept arbitrary system prompts, and readily adopt multiple personae. By analogy to hu…