Nash CoT: Multi-Path Inference with Preference Equilibrium

Open in new window