Explainability on Imagenet
24µFidelity (Uniform)Saliency
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| SaliencyModel Type=OTNN2022.06 | 24 | 14.7 | 0.6 | 0 | |
| Integ. Grad.Model Type=OTNN2022.06 | 4.6 | 5.7 | 0.6 | — | |
| GradCamModel Type=Unconstrained2022.06 | 4.6 | 6.8 | — | — | |
| GradCamModel Type=OTNN2022.06 | 2.9 | 4.8 | — | — | |
| Integ. Grad.Model Type=Unconstrained2022.06 | 2.2 | 2.3 | 0.74 | — | |
| Grad. InputModel Type=OTNN2022.06 | 0.9 | 2 | — | — | |
| SaliencyModel Type=Unconstrained2022.06 | 0.4 | 4.9 | 0.74 | 0 | |
| SmoothGradModel Type=OTNN2022.06 | 0.1 | 0.1 | 0.62 | — | |
| Grad. InputModel Type=Unconstrained2022.06 | 0 | -0.1 | — | — | |
| SmoothGradModel Type=Unconstrained2022.06 | -0.2 | 0.2 | 0.82 | — |