List your company

Model Ranking Correlation Analysis on SEED-Bench (10%-50% subsets)

0.94Spearman Correlation (10% subset)M2IRT
0.39920.53960.680.8204Mar 3, 2026

Evaluation results

MethodLinks
M2IRT2026.030.940.960.970.970.95
M3IRT2026.030.940.950.950.950.95
FlashEval2026.030.890.90.90.90.9
IRT2026.030.860.830.840.810.79
Random2026.030.860.870.890.910.91
TinyBenchmarks2026.030.690.750.770.80.81
MIRT2026.030.420.590.710.770.82