GUT: Quantifying and Optimizing the Reasoning Uncertainty of LLMs via Graph Complexity

The paper introduces GUT, a method that uses directed acyclic graphs to represent all possible reasoning branches of Large Language Models (LLMs). It comprises two modules: GUT-Q, which quantifies reasoning uncertainty by approximating graph complexity, and GUT-O, which reduces uncertainty through reinforcement learning that rewards lower uncertainty. Experiments on four LLMs across five datasets demonstrate GUT’s effectiveness in measuring and mitigating reasoning uncertainty.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

• llms
• reinforcement-learning