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

BabyAI-Text

Benchmarks

Task NameDataset NameSOTA ResultTrend
Object PickupBabyAI-Text PickupLocal (No Change)
Success Rate45
5
Object PickupBabyAI-Text PickupDist (New Object)
Success Rate71
5
Object PickupBabyAI-Text PickupDist (No Change)
Success Rate75
5
Obstacle NavigationBabyAI-Text GoToLocal (New Object)
Success Rate85
5
Obstacle NavigationBabyAI-Text GoToLocal (No Change)
Success Rate91
5
Multi-room ExplorationBabyAI-Text FindObj (No Change)
Success Rate23
4
Object PickupBabyAI-Text PickupLocal New Object
Success Rate52
3
Multi-room ExplorationBabyAI-Text FindObj New Object
Success Rate20
2
Multi-step ReasoningBabyAI-Text UnlockLocal (New Object)
Success Rate95
2
Multi-step ReasoningBabyAI-Text UnlockLocal (No Change)
Success Rate95
2
Showing 10 of 10 rows