English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

In many scenarios, human decisions are explained based on some high-levelconcepts. In this work, we take a step in the interpretability of neuralnetworks by examining their internal representation or neuron's activationsagainst concepts. A concept is characterized by a set of samples that havespecific features in common. We propose a framework to check the existence of acausal relationship between a concept (or its negation) and task classes. Whilethe previous methods focus on the importance of a concept to a task class, wego further and introduce four measures to quantitatively determine the order ofcausality. Moreover, we propose a method for constructing a hierarchy ofconcepts in the form of a concept-based decision tree which can shed light onhow various concepts interact inside a neural network towards predicting outputclasses. Through experiments, we demonstrate the effectiveness of the proposedmethod in explaining the causal relationship between a concept and thepredictive behaviour of a neural network as well as determining theinteractions between different concepts through constructing a concepthierarchy.

Cause and Effect: Hierarchical Concept-based Explanation of Neural Networks