Do Large Language Models Exhibit Cognitive Dissonance? Studying the Difference Between Revealed Beliefs and Stated Answers

Multiple Choice Questions (MCQ) have become a commonly used approach to assess the capabilities of Large Language Models (LLMs), due to their ease of manipulation and evaluation. The experimental appraisals of the LLMs' Stated Answer (their answer to MCQ) have pointed to their apparent ability to perform probabilistic reasoning or to grasp uncertainty. In this work, we investigate whether these aptitudes are measurable outside tailored prompting and MCQ by reformulating these issues as direct text-completion - the fundamental computational unit of LLMs. We introduce Revealed Belief, an evaluation framework that evaluates LLMs on tasks requiring reasoning under uncertainty, which complements MCQ scoring by analyzing text-completion probability distributions. Our findings suggest that while LLMs frequently state the correct answer, their Revealed Belief shows that they often allocate probability mass inconsistently, exhibit systematic biases, and often fail to update their beliefs appropriately when presented with new evidence, leading to strong potential impacts on downstream tasks. These results suggest that common evaluation methods may only provide a partial picture and that more research is needed to assess the extent and nature of their capabilities.