Task Predictive Performance on CelebA (test)
70.02Accuracy/AUC (%)IntCEM
Evaluation Results
| Method | Links | |
|---|---|---|
| IntCEMFraction of concepts intervened=100%, Intervention Policy=Random2023.09 | 70.02 | |
| IntCEMFraction of concepts intervened=100%, Intervention Policy=Learnt Policy ψ2023.09 | 70.02 | |
| IntCEMFraction of concepts intervened=75%, Intervention Policy=Learnt Policy ψ2023.09 | 66.48 | |
| IntCEMFraction of concepts intervened=50%, Intervention Policy=Learnt Policy ψ2023.09 | 62.01 | |
| IntCEMFraction of concepts intervened=75%, Intervention Policy=Random2023.09 | 56.99 | |
| IntCEMFraction of concepts intervened=50%, Intervention Policy=Random2023.09 | 52.01 | |
| IntCEMFraction of concepts intervened=25%, Intervention Policy=Learnt Policy ψ2023.09 | 47.63 | |
| IntCEMFraction of concepts intervened=25%, Intervention Policy=Random2023.09 | 42.25 | |
| IntCEMlambda_roll=52023.09 | 32.07 | |
| IntCEMlambda_roll=12023.09 | 32.05 | |
| IntCEMlambda_roll=0.12023.09 | 31.04 | |
| CEM2023.09 | 31.04 | |
| Seq. CBMTraining paradigm=Sequential2023.09 | 24.53 | |
| Ind. CBMTraining paradigm=Independent2023.09 | 24.25 | |
| Joint CBMBottleneck activation function=Sigmoid2023.09 | 24.15 | |
| Joint CBMBottleneck activation function=Logit2023.09 | 23.87 |