Clarify, Abstain or Answer? Strategising in Conversation with Belief-Augmented Generation

ArXi:2605.25831v1 Announce Type: cross Large language models (LLMs) define a distribution over text, which can be viewed as a probabilistic representation of uncertainty: sampling K responses yields a belief state - responses a model deems plausible. Existing work exploits this representation for narrow tasks like either decoding or selective prediction, and often requires manual interventions, not controlling generation directly.