Towards a Better Understanding of Deep Neural Networks Representations using Deep Generative Networks

Despraz, Jérémie; Gomez, Stéphane; Satizábal, Héctor F.; Peña-Reyes, Carlos Andrés

doi:10.5220/0006495102150222

Cited by 10 publications

(6 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…• Activation maximization, for example, based on Generative Adversarial Networks [56], use deep generative networks and tailored optimization methods to generate class-relevant inputs for convolutional neural networks [13]. A human user can then understand the internal representations assimilated by the network and the typical representations of the classes.…”

Section: Post-hoc Approaches: Explain Machine Learning Modelsmentioning

confidence: 99%

The coming of age of interpretable and explainable machine learning models

Lisboa

Saralajew²,

Vellido

et al. 2023

Neurocomputing

View full text Add to dashboard Cite

Section: Post-hoc Approaches: Explain Machine Learning Modelsmentioning

confidence: 99%

The coming of age of interpretable and explainable machine learning models

Lisboa

Saralajew²,

Vellido

et al. 2023

Neurocomputing

View full text Add to dashboard Cite

“…Attributions or feature attributions are one of the most popular techniques used to explain the model’s predictions . The attribution method assigns scores to each input feature that reflects the contribution of that feature to an ML model’s prediction, thereby explaining the role played by that feature in the prediction. ,, In the case of GNNs, attribution methods assign attribution scores to graph nodes and edges based on their contributions to the final prediction of the model.…”

Section: Methodsmentioning

confidence: 99%

Integrating Explainability into Graph Neural Network Models for the Prediction of X-ray Absorption Spectra

Kotobi,

Singh,

Höche

et al. 2023

J. Am. Chem. Soc.

View full text Add to dashboard Cite

The use of sophisticated machine learning (ML) models, such as graph neural networks (GNNs), to predict complex molecular properties or all kinds of spectra has grown rapidly. However, ensuring the interpretability of these models' predictions remains a challenge. For example, a rigorous understanding of the predicted X-ray absorption spectrum (XAS) generated by such ML models requires an in-depth investigation of the respective black-box ML model used. Here, this is done for different GNNs based on a comprehensive, custom-generated XAS data set for small organic molecules. We show that a thorough analysis of the different ML models with respect to the local and global environments considered in each ML model is essential for the selection of an appropriate ML model that allows a robust XAS prediction. Moreover, we employ feature attribution to determine the respective contributions of various atoms in the molecules to the peaks observed in the XAS spectrum. By comparing this peak assignment to the core and virtual orbitals from the quantum chemical calculations underlying our data set, we demonstrate that it is possible to relate the atomic contributions via these orbitals to the XAS spectrum.

show abstract

“…Another compelling post hoc rationalization is a class of a explanations that reverse-engineer exemplars of a neuron or layer's receptive field. This can be done using activation maximization techniques that find the maximum value of the dot product between an activation vector and some iteratively-sampled image set (Erhan et al, 2009;Yosinski et al, 2015) or an iteratively generated image (Nguyen et al, 2016;Despraz et al, 2017). The maximal argument is taken as that neuron's receptive field, or preferred stimulus.…”

Section: Post Hoc Explanationsmentioning

confidence: 99%

Artificial cognition: How experimental psychology can help generate explainable artificial intelligence

Taylor

2020

Psychon Bull Rev

View full text Add to dashboard Cite

Artificial intelligence powered by deep neural networks has reached a level of complexity where it can be difficult or impossible to express how a model makes its decisions. This black-box problem is especially concerning when the model makes decisions with consequences for human well-being. In response, an emerging field called explainable artificial intelligence (XAI) aims to increase the interpretability, fairness, and transparency of machine learning. In this paper, we describe how cognitive psychologists can make contributions to XAI. The human mind is also a black box, and cognitive psychologists have over one hundred and fifty years of experience modeling it through experimentation. We ought to translate the methods and rigour of cognitive psychology to the study of artificial black boxes in the service of explainability. We provide a review of XAI for psychologists, arguing that current methods possess a blind spot that can be complemented by the experimental cognitive tradition. We also provide a framework for research in XAI, highlight exemplary cases of experimentation within XAI inspired by psychological science, and provide a tutorial on experimenting with machines. We end by noting the advantages of an experimental approach and invite other psychologists to conduct research in this exciting new field.

show abstract

Towards a Better Understanding of Deep Neural Networks Representations using Deep Generative Networks

Cited by 10 publications

References 4 publications

The coming of age of interpretable and explainable machine learning models

The coming of age of interpretable and explainable machine learning models

Integrating Explainability into Graph Neural Network Models for the Prediction of X-ray Absorption Spectra

Artificial cognition: How experimental psychology can help generate explainable artificial intelligence

Contact Info

Product

Resources

About