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

Dialogue Evaluation on SVBench

58.26SA

GPT-4o

36.88842.436547.98553.5335Jun 16, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
58.2664.7670.7567.6855.8262.5761.27
2026.06
56.0362.6169.0965.3653.7360.359.12
2026.06
54.0661.0866.4366.0652.6758.9557.41
2026.06
52.5459.8565.5264.6451.2357.5755.21
2026.06
51.759.565.3361.7250.0956.6353.15
2026.06
47.4956.5863.8659.748.9253.7252.2
2026.06
46.8353.4858.2258.9147.0251.7349.16
2026.06
46.4353.7559.3657.2945.6451.3749.76
2026.06
43.7350.756.6155.0343.4648.7346.89
2026.06
37.7144.5952.0541.836.5841.438.76