[NeurIPS 2025] A Theoretical Study on Bridging Internal Probability and Self-Consistency for LLM Reasoning

๐Ÿ“„ [Paper] โ€ข ๐ŸŒ [Project] โ€ข ๐Ÿ“š [BibTeX]

1. Experimental Settings

Dataset
Model
K (Number of Sampled Reasoning Paths)