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

Four Rooms

Benchmarks

Task NameDataset NameSOTA ResultTrend
Goal-reaching NavigationFour Rooms large-diverse v1 (test)
Success Rate72.2
4
Goal-reaching NavigationFour Rooms large-play v1 (test)
Success Rate67.2
4
Goal-reaching NavigationFour Rooms medium-diverse v1 (test)
Success Rate0.874
4
Goal-reaching NavigationFour Rooms medium-play v1 (test)
Average Success Rate0.787
4
MDP Planning under Reward MisspecificationFour Rooms
Time (msecs)111.53
4
Showing 5 of 5 rows