University of Cambridge > Talks.cam > Artificial Intelligence Research Group Talks (Computer Laboratory) > Global Explainability of GNNs via Logic Combination of Learned Concepts

Global Explainability of GNNs via Logic Combination of Learned Concepts

Add to your list(s) Download to your calendar using vCal

If you have a question about this talk, please contact Pietro Lio.

https://zoom.us/j/99166955895?pwd=SzI0M3pMVEkvNmw3Q0dqNDVRalZvdz09

While instance-level explanation of GNN is a well-studied problem with plenty of approaches being developed, providing a global explanation for the behavior of a GNN is much less explored, despite its potential in interpretability and debugging. Existing solutions either simply list local explanations for a given class, or generate a synthetic prototypical graph with maximal score for a given class, completely missing any combinatorial aspect that the GNN could have learned. In this work, we propose GLG Explainer (Global Logic-based GNN Explainer), the first Global Explainer capable of generating explanations as arbitrary Boolean combinations of learned graphical concepts. GLG Explainer is a fully differentiable architecture that takes local explanations as inputs and combines them into a logic formula over graphical concepts, represented as clusters of local explanations. Contrary to existing solutions, GLG Explainer provides accurate and human-interpretable global explanations that are aligned with ground-truth explanations (on synthetic data) or match existing domain knowledge (on real-world data). Extracted formulas are faithful to the model predictions, to the point of providing insights into some occasionally incorrect rules learned by the model, making GLG Explainer a promising diagnostic tool for learned GNNs.

https://zoom.us/j/99166955895?pwd=SzI0M3pMVEkvNmw3Q0dqNDVRalZvdz09

This talk is part of the Artificial Intelligence Research Group Talks (Computer Laboratory) series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.

 

© 2006-2024 Talks.cam, University of Cambridge. Contact Us | Help and Documentation | Privacy and Publicity