Reward Prediction
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
82.3AlfWorld Reward Score
13
Mar 11, 2026
0.665Overall MAE
12
Feb 26, 2026
4.01MAE
10
Mar 17, 2026
5.45MAE
10
Mar 17, 2026
7.82MAE
10
Mar 17, 2026
170.03MAE
10
Mar 17, 2026
31.91MAE
10
Mar 17, 2026
7.76Mean Absolute Error (MAE)
10
Mar 17, 2026
8.9MAE
10
Mar 17, 2026
5.94MAE
10
Mar 17, 2026
96.5Accuracy
9
Apr 16, 2026
91.4Accuracy
9
Apr 16, 2026
93.1Accuracy
9
Apr 16, 2026
0.006Demo L (MSE)
8
Jun 11, 2026
0.67MAE
8
Mar 4, 2026
30Effort
8
Feb 26, 2026
0.02Demo L (MSE)
7
Jun 11, 2026
0.031Demo L MSE
7
Jun 11, 2026
66.96Average Score
6
Mar 23, 2026
82.22Accuracy
6
Mar 23, 2026
76.89Accuracy
6
Mar 23, 2026
42.22Accuracy
6
Mar 23, 2026
93.5Reward Accuracy
5
Apr 16, 2026
92Reward Accuracy
5
Apr 16, 2026
96Pick Score
5
Mar 6, 2026