Amir Mehrpanah

I'm interested in understanding the underlying dynamics of deep models and doing a PhD on explaining the uncertainty of deep models.

I mostly use theoretical methods to understand deep networks better, see this video for an example. I have got plenty of cool ideas, most of which does not work. But feel free to email me if you are looking for a project about explainability.

Our Recent Publications:

1. On the Complexity-Faithfulness Trade-off of Gradient-Based Explanations [ICCV 25], a short youtube video
2. On Spectral Properties of Gradient-Based Explanation Methods [ECCV 24], a short youtube video

This is my fourth year and I'll soon stop maintaining this page please visit my github.io page