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

CoSPlan

Benchmarks

Task NameDataset NameSOTA ResultTrend
Error DetectionCoSPlan
Robo-VQA-E Score9.1
20
Step CompletionCoSPlan
Robo-VQA-E Step Completion Score56.4
20
Error DetectionCoSPlan Maze-E
Accuracy0.403
15
Step CompletionCoSPlan Blocks-World-E
Accuracy54.3
15
Step CompletionCoSPlan Maze-E
Accuracy46.1
15
Step CompletionCoSPlan Shuffle-E
Accuracy0.301
15
Showing 6 of 6 rows