Layer Wise Relevance Propagation

Layer-wise Relevance Propagation (LRP) is an explainable AI (XAI) technique used to understand the decision-making processes of complex neural networks, particularly deep learning models, by tracing the flow of information backward from the output to the input. Current research focuses on applying LRP to diverse architectures, including transformers, recurrent neural networks (like LSTMs and ESNs), and convolutional neural networks, to improve the interpretability of predictions in various domains such as image classification, natural language processing, and time series analysis. The ability to provide transparent and reliable explanations using LRP is crucial for building trust in AI systems and facilitating their adoption in high-stakes applications where understanding model decisions is paramount.

Papers

October 31, 2024

Zero-shot Class Unlearning via Layer-wise Relevance Analysis and Neuronal Path Perturbation
Wenhan Chang, Tianqing Zhu, Yufeng Wu, Wanlei Zhou
Machine Unlearning Unlearning Framework Layer Wise Relevance Propagation Unlearning Model Zero Shot Unlearning

October 4, 2024

Comparing zero-shot self-explanations with human rationales in multilingual text classification
Stephanie Brandl, Oliver Eberle
Feature Attribution Text Classification Task Sentiment Analysis Task Layer Wise Relevance Propagation Self Explanation Multilingual Text Classification

August 28, 2024

LRP4RAG: Detecting Hallucinations in Retrieval-Augmented Generation via Layer-wise Relevance Propagation
Haichuan Hu, Yuhan Sun, Quanjun Zhang
Large Language Model Retrieval Augmented Generation Layer Wise Relevance Propagation Relevance Modeling Relevance Ranking

August 7, 2024

Methodological Explainability Evaluation of an Interpretable Deep Learning Model for Post-Hepatectomy Liver Failure Prediction Incorporating Counterfactual Explanations and Layerwise Relevance Propagation: A Prospective In Silico Trial
Xian Zhong, Zohaib Salahuddin, Yi Chen, Henry C Woodruff, Haiyi Long, Jianyun Peng, Nuwan Udawatte, Roberto Casale, Ayoub Mokhtari, Xiaoer Zhang, Jiayao Huang, Qingyu Wu, Li Tan, Lili Chen, Dongming Li, Xiaoyan Xie, Manxia Lin, Philippe Lambin
High Explainability Counterfactual Explanation Explainability Method Interpretable Deep Learning Layer Wise Relevance Propagation

July 12, 2024

Layer-Wise Relevance Propagation with Conservation Property for ResNet
Seitaro Otsuki, Tsumugi Iida, Félix Doublet, Tsubasa Hirakawa, Takayoshi Yamashita, Hironobu Fujiyoshi, Komei Sugiura
ResNet Based Layer Wise Relevance Propagation Relevance Score ResNet Model

June 11, 2024

MambaLRP: Explaining Selective State Space Sequence Models
Farnoush Rezaei Jafari, Grégoire Montavon, Klaus-Robert Müller, Oliver Eberle
Layer Wise Relevance Propagation Selective State Space Model

May 10, 2024

Explaining Text Similarity in Transformer Models
Alexandros Vasileiou, Oliver Eberle
Transformer Model Natural Language Processing Model Second Order Semantic Similarity Layer Wise Relevance Propagation Explainable Image Bilinear Model

April 22, 2024

Sparse Explanations of Neural Networks Using Pruned Layer-Wise Relevance Propagation
Paulo Yanez Sarmiento, Simon Witzke, Nadja Klein, Bernhard Y. Renard
Neural Network High Explainability Layer Wise Relevance Propagation

April 16, 2024

Explainable concept mappings of MRI: Revealing the mechanisms underlying deep learning-based brain disease classification
Christian Tinauer, Anna Damulina, Maximilian Sackl, Martin Soellradl, Reduan Achtibat, Maximilian Dreyer, Frederik Pahde, Sebastian Lapuschkin, Reinhold Schmidt, Stefan Ropele, Wojciech Samek, Christian Langkammer
Deep Neural Network Alzheimer'S Disease Magnetic Resonance Imaging Functional Mechanism Layer Wise Relevance Propagation Brain Region

April 4, 2024

Interpreting End-to-End Deep Learning Models for Speech Source Localization Using Layer-wise Relevance Propagation
Luca Comanducci, Fabio Antonacci, Augusto Sarti
End to End Layer Wise Relevance Propagation Relevance Score Speaker Localization

March 14, 2024

EventRPG: Event Data Augmentation with Relevance Propagation Guidance
Mingyuan Sun, Donghao Zhang, Zongyuan Ge, Jiaxu Wang, Jia Li, Zheng Fang, Renjing Xu
Data Augmentation Spiking Neural Network Saliency Map Layer Wise Relevance Propagation Efficient Augmentation

February 8, 2024

AttnLRP: Attention-Aware Layer-Wise Relevance Propagation for Transformers
Reduan Achtibat, Sayed Mohammad Vakilzadeh Hatefi, Maximilian Dreyer, Aakriti Jain, Thomas Wiegand, Sebastian Lapuschkin, Wojciech Samek
Vision Transformer Transformer Megatron Decepticons Latent Representation Attention Layer Layer Wise Relevance Propagation

January 20, 2024

LRP-QViT: Mixed-Precision Vision Transformer Quantization via Layer-wise Relevance Propagation
Navin Ranjan, Andreas Savakis
Vision Transformer Mixed Precision Quantization Layer Wise Relevance Propagation Vision Transformer Quantization Bit Width Quantization

January 18, 2024

Better Explain Transformers by Illuminating Important Information
Linxin Song, Yan Cui, Ao Luo, Freddy Lecue, Irene Li
Transformer Based Model Evaluation Metric Full Information Layer Wise Relevance Propagation

January 16, 2024

Faster ISNet for Background Bias Mitigation on Deep Neural Networks
Pedro R. A. S. Bassi, Sergio Decherchi, Andrea Cavalli
Deep Neural Network Shortcut Learning Layer Wise Relevance Propagation Background Bias

October 31, 2023

Unveiling Black-boxes: Explainable Deep Learning Models for Patent Classification
Md Shajalal, Sebastian Denef, Md. Rezaul Karim, Alexander Boden, Gunnar Stevens
Black Box Layer Wise Relevance Propagation Explainable Deep Learning Patent Classification

October 11, 2023

Human-Centered Evaluation of XAI Methods
Karam Dawoud, Wojciech Samek, Peter Eisert, Sebastian Lapuschkin, Sebastian Bosse
Deep Learning Inherent Interpretability Image Classification XAI Method Layer Wise Relevance Propagation Human Centered Evaluation

October 10, 2023

XAI for Early Crop Classification
Ayshah Chan, Maja Schneider, Marco Körner
xAI Community Layer Wise Relevance Propagation Crop Classification Early Classification

August 23, 2023

Layer-wise Feedback Propagation
Leander Weber, Jim Berend, Alexander Binder, Thomas Wiegand, Wojciech Samek, Sebastian Lapuschkin
Transfer Learning Spiking Neural Network Layer Wise Relevance Propagation Feedback Network Neural Predictor

May 31, 2023

Graph-based methods coupled with specific distributional distances for adversarial attack detection
Dwight Nwaigwe, Lucrezia Carboni, Martial Mermillod, Sophie Achard, Michel Dojat
Neural Network Adversarial Attack Sparse Graph Layer Wise Relevance Propagation Graph Analysis Adversarial Attack Detection Graph Based Method Distance Distribution

Layer Wise Relevance Propagation

Papers

Zero-shot Class Unlearning via Layer-wise Relevance Analysis and Neuronal Path Perturbation

Comparing zero-shot self-explanations with human rationales in multilingual text classification

LRP4RAG: Detecting Hallucinations in Retrieval-Augmented Generation via Layer-wise Relevance Propagation

Methodological Explainability Evaluation of an Interpretable Deep Learning Model for Post-Hepatectomy Liver Failure Prediction Incorporating Counterfactual Explanations and Layerwise Relevance Propagation: A Prospective In Silico Trial

Layer-Wise Relevance Propagation with Conservation Property for ResNet

MambaLRP: Explaining Selective State Space Sequence Models

Explaining Text Similarity in Transformer Models

Sparse Explanations of Neural Networks Using Pruned Layer-Wise Relevance Propagation

Explainable concept mappings of MRI: Revealing the mechanisms underlying deep learning-based brain disease classification

Interpreting End-to-End Deep Learning Models for Speech Source Localization Using Layer-wise Relevance Propagation

EventRPG: Event Data Augmentation with Relevance Propagation Guidance

AttnLRP: Attention-Aware Layer-Wise Relevance Propagation for Transformers

LRP-QViT: Mixed-Precision Vision Transformer Quantization via Layer-wise Relevance Propagation

Better Explain Transformers by Illuminating Important Information

Faster ISNet for Background Bias Mitigation on Deep Neural Networks

Unveiling Black-boxes: Explainable Deep Learning Models for Patent Classification

Human-Centered Evaluation of XAI Methods

XAI for Early Crop Classification

Layer-wise Feedback Propagation

Graph-based methods coupled with specific distributional distances for adversarial attack detection