1 paper · 1 filter
Yanxiao Liu, Sicheng Wan, Deniz Gündüz
Best-of-n (BoN) sampling is a simple yet effective inference-time alignment method, but hard maximization provides only coarse control over the trade-off between reward and distr…