#Alignment

Best-of-Tails: Analyzing Inference-Time Alignment in Large Language Models