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

Task Success Metrics (Real, WM, W2R, CR) on ALFWorld

91Real Success

GPT-5

4.357626.851349.34571.8387Dec 21, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.12
9184.6286.670.95
2025.12
9182.5681.540.9
2025.12
8276760.93
2025.12
8284841.02
2025.12
74.2162.5664.620.87
2025.12
74.2167.5367.010.9
2025.12
67.268.5669.591.04
2025.12
67.268.7268.211.02
2025.12
5855.957.440.99
2025.12
5858.4656.920.98
2025.12
50.551.7952.311.04
2025.12
50.553.3353.331.06
2025.12
7.697.697.691
2025.12
7.699.749.741.27