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

Response Similarity on MT-Bench

94.1Response Similarity

SOMA

38.25252.75167.2581.749May 11, 2026
Updated 2mo ago

Evaluation Results

MethodLinks
2026.05
94.1
2026.05
93.2
2026.05
92.4
2026.05
87.2
2026.05
85.3
2026.05
77.5
2026.05
69.2
2026.05
68
2026.05
62.7
2026.05
53.6
2026.05
51.7
2026.05
40.4