Anna Hedström

Two new approaches enhance reliability in assessing AI model explanations.

A new method to assess neuron explanations in deep learning models.

Exploring evaluation issues in Explainable Artificial Intelligence and the quest for trust.