Maximilian Dreyer

Prototypical Concept-based Explanations, accepted at SAIAD workshop at CVPR 2024.

Repository for "From What to How: Attributing CLIP's Latent Components Reveals Unexpected Semantic Reliance"

Repository for PURE: Turning Polysemantic Neurons Into Pure Features by Identifying Relevant Circuits, accepted at CVPR 2024 XAI4CV Workshop (spotlight)

Repository with files and scripts for the human study of the CRP (Concept-Relevance-Propagation) paper

Concept Relevance Propagation for Localization Models, accepted at SAIAD workshop at CVPR 2023.

Reveal to Revise: An Explainable AI Life Cycle for Iterative Bias Correction of Deep Models. Paper presented at MICCAI 2023 conference.

automatic evaluation of concept example approaches

Benchmarking Generalized Out-of-Distribution Detection

Segmentation models with pretrained backbones. PyTorch.

An eXplainable AI toolkit with Concept Relevance Propagation and Relevance Maximization

Towards Automatic Concept-based Explanations

Code for the Proceedings of the National Academy of Sciences 2020 article, "Understanding the Role of Individual Units in a Deep Neural Network"

PyTorch implementation of DeepLabV3, trained on the Cityscapes dataset.

No description provided.

A set of examples around pytorch in Vision, Text, Reinforcement Learning, etc.

Zennit is a high-level framework in Python using PyTorch for explaining/exploring neural networks using attribution methods like LRP.