Do-Undo
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Do-Undo zero-shot 1.0
8.44Instruction Adherence (IF)
10
Do-Undo FULL (test)
0.882DINO-R
10
Do-Undo (evaluation)
89DINO-R
6
Do-Undo Forward
8.41Instruction Following (IF)
5
Loading the SOTA2 catalog…