Maze Navigation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
0.8Success Rate
25
Feb 26, 2026
97.66Accuracy
18
May 20, 2026
0.996Accuracy
14
Feb 26, 2026
100RSR
11
Jul 7, 2026
99.6Success Rate
11
Mar 4, 2026
100Safety Rate
10
Apr 28, 2026
0.057D_policy
9
Jul 7, 2026
52.6Best Success Rate @ 3
8
May 22, 2026
72.8Success Rate
7
Feb 26, 2026
90.8Success Rate
7
Feb 26, 2026
3,200EM
7
Feb 26, 2026
4,000EM
7
Feb 26, 2026
4,200Exact Match
7
Feb 26, 2026
4,400EM
7
Feb 26, 2026
80EM
7
Feb 26, 2026
92EM
7
Feb 26, 2026
96EM
7
Feb 26, 2026
100Success Rate (SR)
6
Jul 2, 2026
100Success Rate (SR)
6
Jul 2, 2026
100Success Rate (Case 2 Task 2)
6
Jul 2, 2026
100Success Rate
6
Jul 2, 2026
100Success Rate
6
Jul 2, 2026
90Success Rate (SR)
6
Jul 2, 2026
39.6Mean Completion Time
6
Feb 26, 2026
11Mean Completion Time
6
Feb 26, 2026