Relative Attributing Propagation: Interpreting the Comparative Contributions of Individual Units in Deep Neural Networks

Nam, Woo-Jeoung; Gur, Shir; Choi, Jaesik; Wolf, Lior; Lee, Seong Whan

doi:10.1609/aaai.v34i03.5632

Cited by 60 publications

(37 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The space may be analyzed and visualized to help users understand the performance of these combinations, with respect to image classes and CNN models. This is out of the scope of VisLRPDesigner as an LRP design tool but leads to a critical direction in future work. LRP for DNNs : More technique options could be added in the future, including layers fusing for Batch‐Normalization [GHK ∗ 20] vs. its bypassing in this system, bias switching, and different attribute‐discriminative LRP approaches [NGC ∗ 20]. Besides, While designed mostly for CNNs, LRP is also used in other DNNs such as natural language processing [AHM ∗ 16], EEG analysis [SLSM16], and audio classification [BAL ∗ 18].…”

Section: Discussion and Future Workmentioning

confidence: 99%

A Visual Designer of Layer‐wise Relevance Propagation Models

Huang

Jamonnak

Zhao

et al. 2021

Computer Graphics Forum

View full text Add to dashboard Cite

Layer‐wise Relevance Propagation (LRP) is an emerging and widely‐used method for interpreting the prediction results of convolutional neural networks (CNN). LRP developers often select and employ different relevance backpropagation rules and parameters, to compute relevance scores on input images. However, there exists no obvious solution to define a “best” LRP model. A satisfied model is highly reliant on pertinent images and designers' goals. We develop a visual model designer, named as VisLRPDesigner, to overcome the challenges in the design and use of LRP models. Various LRP rules are unified into an integrated framework with an intuitive workflow of parameter setup. VisLRPDesigner thus allows users to interactively configure and compare LRP models. It also facilitates relevance‐based visual analysis with two important functions: relevance‐based pixel flipping and neuron ablation. Several use cases illustrate the benefits of VisLRPDesigner. The usability and limitation of the visual designer is evaluated by LRP users.

show abstract

Section: Discussion and Future Workmentioning

confidence: 99%

A Visual Designer of Layer‐wise Relevance Propagation Models

Huang

Jamonnak

Zhao

et al. 2021

Computer Graphics Forum

View full text Add to dashboard Cite

show abstract

“…In this section, I have chosen some representative examples to illustrate the application of my proposed taxonomy in Figure 4. Within the category of Diagnostic-explanation are Saliency Maps [25], LIME [29], and Shapley Values [13] which identify particular input features important to affecting the output of models. The Explication-explanation category focuses on techniques to render explanations or model output understandable to humans interacting with the AI model.…”

Section: Organizing Present Xai Methodsmentioning

confidence: 99%

Explanatory Pluralism in Explainable AI

Yao

2021

Preprint

View full text Add to dashboard Cite

The increasingly widespread application of AI models motivates increased demand for explanations from a variety of stakeholders. However, this demand is ambiguous because there are many types of 'explanation' with different evaluative criteria. In the spirit of pluralism, I chart a taxonomy of types of explanation and the associated XAI methods that can address them. When we look to expose the inner mechanisms of AI models, we develop Diagnostic-explanations. When we seek to render model output understandable, we produce Explicationexplanations. When we wish to form stable generalizations of our models, we produce Expectation-explanations. Finally, when we want to justify the usage of a model, we produce Role-explanations that situate models within their social context. The motivation for such a pluralistic view stems from a consideration of causes as manipulable relationships and the different types of explanations as identifying the relevant points in AI systems we can intervene upon to affect our desired changes. This paper reduces the ambiguity in use of the word 'explanation' in the field of XAI, allowing practitioners and stakeholders a useful template for avoiding equivocation and evaluating XAI methods and putative explanations.

show abstract

“…More specifically, we use LRP to calculate relevance heatmaps which highlight regions in input space that speak for or against the classification decision. The LRP-implementations follow [34] for ResNet and [35], [36] for transformers. We implemented LRP rules for the BoTNet, the Inception model as well as the hybrid models along the same lines, none of which have been discussed in the literature so far.…”

Section: B Modelsmentioning

confidence: 99%

From CNNs to Vision Transformers -- A Comprehensive Evaluation of Deep Learning Models for Histopathology

Springenberg¹,

Frommholz²,

Wenzel³

et al. 2022

Preprint

View full text Add to dashboard Cite

While machine learning is currently transforming the field of histopathology, the domain lacks a comprehensive evaluation of state-of-the-art models based on essential but complementary quality requirements beyond a mere classification accuracy. In order to fill this gap, we conducted an extensive evaluation by benchmarking a wide range of classification models, including recent vision transformers, convolutional neural networks and hybrid models comprising transformer and convolutional models. We thoroughly tested the models on five widely used histopathology datasets containing whole slide images of breast, gastric, and colorectal cancer and developed a novel approach using an image-to-image translation model to assess the robustness of a cancer classification model against stain variations. Further, we extended existing interpretability methods to previously unstudied models and systematically reveal insights of the models' classification strategies that allow for plausibility checks and systematic comparisons. The study resulted in specific model recommendations for practitioners as well as putting forward a general methodology to quantify a model's quality according to complementary requirements that can be transferred to future model architectures.

show abstract

Relative Attributing Propagation: Interpreting the Comparative Contributions of Individual Units in Deep Neural Networks

Cited by 60 publications

References 16 publications

A Visual Designer of Layer‐wise Relevance Propagation Models

A Visual Designer of Layer‐wise Relevance Propagation Models

Explanatory Pluralism in Explainable AI

From CNNs to Vision Transformers -- A Comprehensive Evaluation of Deep Learning Models for Histopathology

Contact Info

Product

Resources

About