Bridging Internal Probability and Self-Consistency for Effective and Efficient LLM Reasoning

Open in new window