Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Pairwise Comparison on AlignBench

74.69Agreement

GPT-4

15.472430.846246.2261.5938Nov 30, 2023
Updated 4d ago

Evaluation Results

MethodLinks
2023.11
74.6986.75
2023.11
70.5689.25
2023.11
70.2584.88
2023.11
61.6974.06
2023.11
58.8183.06
2023.11
53.8872.25
2023.11
42.566
2023.11
42.0658.75
2023.11
41.3864.19
2023.11
40.5657.13
2023.11
39.5653.94
2023.11
35.8150.06
2023.11
33.8143.25
2023.11
32.538.56
2023.11
27.0640.82
2023.11
2645.38
2023.11
24.7542.88
2023.11
17.7531.84