Large Language Models as Nondeterministic Causal Models

ArXi:2509.22297v2 Announce Type: replace Recent work by Chatzi and Ravfogel has developed, for the first time, a method for generating counterfactuals of probabilistic Large Language Models. Such counterfactuals tell us what would - or might - have been the output of an LLM if some factual prompt ${\bf x}$ had been ${\bf x}^*$ instead. The ability to generate such counterfactuals is an important necessary step towards explaining, evaluating, and eventually improving, the behavior of LLMs.