puzzle
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
puzzle 4x5 play v0
18Success Rate
6
puzzle 4x4 play v0
88Average Binary Success Rate
6
PUZZLE Avg across difficulty levels
54.3Accuracy
4
PUZZLE Human hard
5.7Accuracy
4
PUZZLE Human normal
58.1Accuracy
4
PUZZLE (Human easy)
69.5Accuracy
4
PUZZLE (Trivial)
83.8Accuracy
4