List your company

ResearchFormerly WizWand

TOM-SB

Benchmarks

TaskDatasetSOTA resultTrendResults

Adversarial Theory of Mind

TOM-SB

42.4Fooling Rate (Hard)

9

TOM-SB

72ToM Accuracy (Trajectory)

3

Adversarial Defense

TOM-SB

11.5Fooling % (Hard)

3