MAZE
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Maze Hard (test)
87.7Accuracy
3
Maze-hard (test)
88.8FSR
3
Maze 48x48
89Validity
3
Maze 32x32
88.6Validity
3
Maze 16x16
88.6Validity
3
Maze 8x8
94Validity
3
Maze 17^10 hybrid
9.72Mean Score
3
Maze 5^4 hybrid
9.78Mean Score
3
MAZE Ant Room5x11 Robustness evaluation v1
0.332Robustness Rate
3
MAZE Ant UMaze Robustness evaluation v1
64.8Robustness Rate
3
MAZE Point Corridor7x7 v1 (Robustness evaluation)
38.9Robustness (%)
3
MAZE Point Room5x11 Robustness evaluation v1
38.6Robustness Rate
3
Maze
244.291Time (msecs)
2
Maze
82.2Accuracy
1
Maze
—Primary metric
0