Best-of-N Sampling
Also called rejection sampling, best-of-n, BoN.
Best-of-N sampling generates N candidate outputs and returns the one ranked highest by a verifier, reward model, or scoring function.
Description
It is the simplest form of trading inference compute for quality. Its effectiveness depends on the verifier, and optimizing too hard against an imperfect reward model can select outputs that game it.
Sources
- Lightman et al. (2023). Let's Verify Step by Step.
- Snell et al. (2024). Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters.
Cite this entry
Protologue. (2026). Best-of-N Sampling. In Protologue: A Taxonomy of Prompting and LLM Techniques (v1.0.0, PTL-0054). https://protologue.com/t/best-of-n-sampling/
BibTeX
@misc{protologue_best_of_n_sampling,
title = {Best-of-N Sampling},
author = {{Protologue}},
year = {2026},
howpublished = {Protologue: A Taxonomy of Prompting and LLM Techniques, v1.0.0},
note = {Entry PTL-0054},
url = {https://protologue.com/t/best-of-n-sampling/}
}