Smoothed Best-of-N has finite-sample KL and regret bounds under imperfect reward models, and tuning its temperature can make its regret bound beat hard Best-of-N in the overoptimization regime.
Variational best-of-n alignment
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
method 1
citation-polarity summary
fields
stat.ML 1years
2025 1verdicts
CONDITIONAL 1roles
method 1polarities
use method 1representative citing papers
citing papers explorer
-
Best-of-N through the Smoothing Lens: KL Divergence and Regret Analysis
Smoothed Best-of-N has finite-sample KL and regret bounds under imperfect reward models, and tuning its temperature can make its regret bound beat hard Best-of-N in the overoptimization regime.