AI Enhancing Large Language Models’ Reflection: Tackling Overconfidence and Randomness with Self-Contrast for Improved Stability and Accuracy
