Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

SCOUT

Benchmarks

Task NameDataset NameSOTA ResultTrend
Prompt Injection DetectionSCOUT-450
ASR (hid)0
13
Prompt injection detectionSCOUT-450 (Held-out evaluation)
Accuracy92.4
5
Traversable-Obstacle NavigationSCOUT Real-world 2.0
Success Rate80
4
Combined Task NavigationReal-world SCOUT 2.0 platform
Success Rate (SR)90
3
Showing 4 of 4 rows