Adam Block

Papers on Lattice

Total citations

Topics

Research focus

Constitutional AI & AI Ethics (1)Red-Teaming & Adversarial Robustness (1)RLHF & Preference Learning (1)

Frequent co-authors

Zhuohao Yu (1)Zhiwei Steven Wu (1)

Papers (1)

Apr 6, 2026

Zhuohao Yu +2Apr 6, 2026·also CMU ML

From Curiosity to Caution: Mitigating Reward Hacking for Best-of-N with Pessimism

Counterintuitively, penalizing prediction error in language model rewards—instead of rewarding it—can drastically improve Best-of-N sampling by preventing reward hacking.

Zhuohao Yu, Zhiwei Steven Wu, Adam Block

Constitutional AI & AI Ethics Red-Teaming & Adversarial Robustness RLHF & Preference Learning

Search

Adam Block

Research focus

Frequent co-authors

Papers (1)