Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Pairwise Preference Comparison research benchmarks · SOTA2 Research

Pairwise Preference Comparison

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

Pairwise Preference Comparisons 1.5B Scale (test)

STACKELBERGGDA FOLLOWER

0.834Avg Preference Score

30

Feb 26, 2026

Qwen2.5-3B responses (test)

82.7Avg Preference Score

30

Feb 26, 2026

150 prompt-response pairs

RL-trained policy

63.333Win Rate

9

May 29, 2026

HH-RLHF held-out (test)

53.02Win Rate

6

Mar 25, 2026

Popular datasets

HH-RLHF Qwen2.5 Pairwise Preference Comparisons prompt-response pairs

Follow for updates

Join SOTA2 Contact