Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Decision-making in interactive environments on ScienceWorld (OOD)
Loading...
34.36
Performance
Glider
3.42
11.4525
19.485
27.5175
Jun 2, 2026
Performance
Avg Token Usage
Normalized Token Efficiency (TE)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Performance
Avg Token Usage
Normalized Token Efficiency (TE)
Glider
Training Type=Fine-Tun...
2026.06
34.36
20,939.88
10.838
Multi2
Training Type=Fine-Tun...
2026.06
30.68
14,716.87
13.769
ADaPT
Training Type=Prompt-B...
2026.06
12.27
63,288.85
1.281
ReAct
Training Type=Prompt-B...
2026.06
10.3
68,031.35
1
Reflexion
Training Type=Prompt-B...
2026.06
6.97
64,755.37
0.711
GRPO
Training Type=Fine-Tun...
2026.06
4.61
9,878
3.083
Feedback
Search any
task
Search any
task