Is Best-of-N the Best of Them? Coverage, Scaling, and Optimality in Inference-Time Alignment

Audrey Huang; Adam Block; Qinghua Liu; Nan Jiang; Akshay Krishnamurthy; Dylan J Foster

2025 ICML ICML 2025

Is Best-of-N the Best of Them? Coverage, Scaling, and Optimality in Inference-Time Alignment

Abstract

No abstract available. No abstract was captured for this paper when its proceedings were scraped - most often an older edition published without one online. How the corpus is built.

❓ The Questioner

Authors

Audrey Huang, Adam Block, Qinghua Liu, Nan Jiang, Akshay Krishnamurthy, Dylan J Foster

Topics

Artificial Intelligence > Core AI > AI Safety Machine Learning > Optimization & Theory > Optimization Artificial Intelligence > Core AI > Large Language Models

Keywords

scalability reward model best-of-n sampling inference-time alignment pessimism coverage

Download PDF