List your company

Metric Correlation with Human Judgment on Merlin

0.369Pearson CorrelationRadGraph
0.11420.180350.24650.31265Apr 27, 2026

Evaluation results

MethodLinks
RadGraph2026.040.3690.2510.341
RaTEScore2026.040.3340.2270.303
CT-FineBench2026.040.3260.3060.401
BERTScore2026.040.3110.2170.293
GREEN2026.040.3090.1790.228
BLEU-22026.040.1680.0890.114
ROUGE-L2026.040.1240.070.097