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

Next-state prediction on TextWorld (TW)

70.6EM Accuracy

Qwen2.5-7B

-2.82416.23835.354.362Dec 21, 2025
Updated 2d ago

Evaluation Results

MethodLinks
2025.12
70.6
2025.12
70.45
2025.12
49.12
2025.12
44.27
2025.12
40.35
2025.12
17.7
2025.12
14.11
2025.12
13.39
2025.12
11.66
2025.12
11.43
2025.12
9.2
2025.12
7.86
2025.12
3.51
2025.12
0.36
2025.12
0
2025.12
0