Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Goal Prediction on Maze OOD
Loading...
70
Classification Accuracy
SVoTp
28.4
39.2
50
60.8
Jun 10, 2026
Classification Accuracy
Free-response Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Classification Accuracy
Free-response Accuracy
SVoTp
Size=5, Backbone=Anole
2026.06
70
41.7
SVoTp
Size=6, Backbone=Anole
2026.06
66.7
36.7
SVoTp
Size=7, Backbone=Anole
2026.06
65
33.3
SVoTo
Size=5, Backbone=Anole
2026.06
60
33.3
T-CoT
Size=5, Backbone=GPT-4o
2026.06
56.7
26.7
MVoT
Size=5, Backbone=Anole
2026.06
56.7
26.7
SVoTo
Size=6, Backbone=Anole
2026.06
55
26.7
SVoTo
Size=7, Backbone=Anole
2026.06
51.7
20
T-CoT
Size=5, Backbone=Anole
2026.06
46.7
16.7
MVoT
Size=6, Backbone=Anole
2026.06
46.7
3.3
T-CoT
Size=6, Backbone=GPT-4o
2026.06
40
16.7
MVoT
Size=7, Backbone=Anole
2026.06
40
0
T-CoT
Size=6, Backbone=Anole
2026.06
36.7
1.7
T-CoT
Size=7, Backbone=GPT-4o
2026.06
36.7
10
T-CoT
Size=7, Backbone=Anole
2026.06
30
0
Feedback
Search any
task
Search any
task