activity
20202022
most citedGCExplainer: Human-in-the-Loop Concept-based Explanations for Graph Neural Networks

14 citations · 17 across the 5 of their papers we have counts for

collaborators
Showing cs.LGShow all

9 papers · 1 filter

cs.LG20221 cited

Explainer Divergence Scores (EDS): Some Post-Hoc Explanations May be Effective for Detecting Unknown Spurious Correlations

Shea Cardozo, Gabriel Islas Montero, Dmitry Kazhdan +4

Recent work has suggested post-hoc explainers might be ineffective for detecting spurious correlations in Deep Neural Networks (DNNs). However, we show there are serious weaknesses…

cs.LG202114 cited

GCExplainer: Human-in-the-Loop Concept-based Explanations for Graph Neural Networks

Lucie Charlotte Magister, Dmitry Kazhdan, Vikash Singh +1

While graph neural networks (GNNs) have been shown to perform well on graph-based data from a variety of fields, they suffer from a lack of transparency and accountability, which h…

cs.LG2021

Algorithmic Concept-based Explainable Reasoning

Dobrik Georgiev, Pietro Barbiero, Dmitry Kazhdan +2

Recent research on graph neural network (GNN) models successfully applied GNNs to classical graph algorithms and combinatorial optimisation problems. This has numerous benefits, su…

cs.LG2021

Failing Conceptually: Concept-Based Explanations of Dataset Shift

Maleakhi A. Wijaya, Dmitry Kazhdan, Botty Dimanov +1

Despite their remarkable performance on a wide range of visual tasks, machine learning technologies often succumb to data distribution shifts. Consequently, a range of recent work…

cs.LG20211 cited

Is Disentanglement all you need? Comparing Concept-based & Disentanglement Approaches

Dmitry Kazhdan, Botty Dimanov, Helena Andres Terre +3

Concept-based explanations have emerged as a popular way of extracting human-interpretable representations from deep discriminative models. At the same time, the disentanglement le…

cs.LG2021

Manipulating SGD with Data Ordering Attacks

Ilia Shumailov, Zakhar Shumaylov, Dmitry Kazhdan +4

Machine learning is vulnerable to a wide variety of attacks. It is now well understood that by changing the underlying data distribution, an adversary can poison the model trained…