Global ETD Search

Help Wanted: Robust Concept Interventions for Interpretable Deep Neural Networks

Espinosa Zarlenga, MateoUniversity of Cambridge2025

… support human-in-the-loop feedback via so-called concept interventions, which are operations where an expert communicates the presence or absence of a high-level concept to the model through a direct manipulation of its latent space. These interventions provide a practical approach to deploying …

cambridge Repository record for Help Wanted: Robust Concept Interventions for Interpretable Deep Neural Networks (opens in a new tab)