List your company

Model Ranking Prediction on Helpsteer 13B+ Models Holdout (test)

74.1Acc_pair (RM1 Helpful)BENCHALIGN
55.06860.00964.9569.891Feb 2, 2026

Evaluation results

MethodLinks
BENCHALIGNHoldout=13B+ Models2026.0274.169.60.6740.552
METABENCHHoldout=13B+ Models2026.0258.4530.2430.089
TINYBENCHMARKSHoldout=13B+ Models2026.0257520.2030.061
RANDOMHoldout=13B+ Models2026.0255.853.70.1250.083