14 papers
Are Safety Guarantees in Neural Networks Safe? How to Compute Trustworthy Robustness Certifications
Merkouris Papamichail, Konstantinos Varsos, Giorgos Flouris +1
A primary challenge in AI safety is the existence of adversarial examples -- slightly distorted inputs that cause a neural network (NN) to misclassify. To mitigate this problem, re…
Towards Rigorous Explainability by Feature Attribution
Olivier Létoffé, Xuanxiang Huang, Joao Marques-Silva
For around a decade, non-symbolic methods have been the option of choice when explaining complex machine learning (ML) models. Unfortunately, such methods lack rigor and can mislea…
The Cost of Relaxation: Evaluating the Error in Convex Neural Network Verification
Merkouris Papamichail, Konstantinos Varsos, Giorgos Flouris +1
Many neural network (NN) verification systems represent the network's input-output relation as a constraint program. Sound and complete, representations involve integer constraints…
Interval Certifications for Multilayered Perceptrons via Lattice Traversal
Merkouris Papamichail, Konstantinos Varsos, Giorgos Flouris +1
In this work we present a rigorous theoretical framework to a foundational problem of AI safety, namely adversarial robustness. In particular, we show that the adversarial robustne…
Uncovering Bugs in Formal Explainers: A Case Study with PyXAI
Xuanxiang Huang, Yacine Izza, Alexey Ignatiev +1
Formal explainable artificial intelligence (XAI) offers unique theoretical guarantees of rigor when compared to other non-formal methods of explainability. However, little attentio…
Efficient & Correct Predictive Equivalence for Decision Trees
Joao Marques-Silva, Alexey Ignatiev
The Rashomon set of decision trees (DTs) finds importance uses. Recent work showed that DTs computing the same classification function, i.e. predictive equivalent DTs, can represen…