List your company

Earth Science Reasoning on EarthSE (test)

58Initial ScoreQWEN3-MAX
38.03243.21648.453.584May 11, 2026

Evaluation results

MethodLinks
QWEN3-MAXModel Scale=Max2026.05580.821.258.8
SFTModel Scale=9B2026.0548.801.61.648.8
CRITIQUEGRPOModel Scale=9B2026.054826450
GRPOModel Scale=9B2026.0546.81.21.60.448
ReCritModel Scale=9B2026.0546.49.613.6456
DPOModel Scale=9B2026.0545.60.82.8246.4
BASE MODELModel Scale=9B2026.0544.40.43.63.244.8
DPOModel Scale=4B2026.0541.2-1.20.41.640
CRITIQUEGRPOModel Scale=4B2026.0540.83.65.21.644.4
SFTModel Scale=4B2026.0540.4-0.80.81.639.6
GRPOModel Scale=4B2026.0540-0.81.2239.2
BASE MODELModel Scale=4B2026.0539.2-1.601.637.6
ReCritModel Scale=4B2026.0538.819.222.83.658