Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Response Quality Evaluation on Held-out n=300 (test)

0.747Pearson Correlation Coefficient

DeBERTa Judge

-0.4074-0.10770.1920.4917Apr 20, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.04
0.7470.6630.733
2026.04
0.6760.5670.685
2026.04
0.6470.5780.698
2026.04
0.6290.5540.683
2026.04
0.6290.5580.7
2026.04
0.4720.3460.452
2026.04
0.3310.2380.357
2026.04
-0.3630.48-0.25