Delivering inflated explanations
In the quest for Explainable Artificial Intelligence (XAI) one of the questions that frequently
arises given a decision made by an AI system is,``why was the decision made in this …
arises given a decision made by an AI system is,``why was the decision made in this …
Logic-based explainability: past, present and future
J Marques-Silva - International Symposium on Leveraging Applications of …, 2024 - Springer
In recent years, the impact of machine learning (ML) and artificial intelligence (AI) in society
has been absolutely remarkable. This impact is expected to continue in the foreseeable …
has been absolutely remarkable. This impact is expected to continue in the foreseeable …
Anytime approximate formal feature attribution
Widespread use of artificial intelligence (AI) algorithms and machine learning (ML) models
on the one hand and a number of crucial issues pertaining to them warrant the need for …
on the one hand and a number of crucial issues pertaining to them warrant the need for …
Computing inflated explanations for boosted trees: a compilation-based approach
Explaining a classification made by tree-ensembles is an inherently hard problem that is
traditionally solved approximately, without guaranteeing sufficiency or necessity. Abductive …
traditionally solved approximately, without guaranteeing sufficiency or necessity. Abductive …
Efficient Contrastive Explanations on Demand
Y Izza, J Marques-Silva - arXiv preprint arXiv:2412.18262, 2024 - arxiv.org
Recent work revealed a tight connection between adversarial robustness and restricted
forms of symbolic explanations, namely distance-based (formal) explanations. This …
forms of symbolic explanations, namely distance-based (formal) explanations. This …
From SHAP Scores to Feature Importance Scores
A central goal of eXplainable Artificial Intelligence (XAI) is to assign relative importance to
the features of a Machine Learning (ML) model given some prediction. The importance of …
the features of a Machine Learning (ML) model given some prediction. The importance of …
The Explanation Game--Rekindled (Extended Version)
J Marques-Silva, X Huang, O Letoffe - arXiv preprint arXiv:2501.11429, 2025 - arxiv.org
Recent work demonstrated the existence of critical flaws in the current use of Shapley values
in explainable AI (XAI), ie the so-called SHAP scores. These flaws are significant in that the …
in explainable AI (XAI), ie the so-called SHAP scores. These flaws are significant in that the …
The Sets of Power
J Marques-Silva, C Mencía, R Mencía - arXiv preprint arXiv:2410.07867, 2024 - arxiv.org
Measures of voting power have been the subject of extensive research since the mid 1940s.
More recently, similar measures of relative importance have been studied in other domains …
More recently, similar measures of relative importance have been studied in other domains …
CAGE: Causality-Aware Shapley Value for Global Explanations
Abstract As Artificial Intelligence (AI) is having more influence on our everyday lives, it
becomes important that AI-based decisions are transparent and explainable. As a …
becomes important that AI-based decisions are transparent and explainable. As a …