Adrien Bennetot

Papers

Total Citations

H-Index

About

Adrien Bennetot is a researcher at the frontier of explainable artificial intelligence (XAI) and human-robot collaboration, with a focus on making machine learning systems more transparent and intuitive. His work centers on understanding how artificial agents construct knowledge internally, drawing inspiration from human cognitive development. Bennetot’s most cited paper, “Explaining Aha! moments in artificial agents through IKE-XAI: Implicit Knowledge Extraction for eXplainable AI” (2022, 20 citations), introduces a novel framework that analyzes how an agent’s latent representations evolve during learning—mirroring the sudden insights, or “Aha! moments,” seen in human problem-solving. This contribution bridges cognitive science and AI, offering tools to visualize and interpret otherwise opaque neural network reasoning. In related work, “Should artificial agents ask for help in human-robot collaborative problem-solving?” (2020, 5 citations), he explores how robots can strategically request assistance from humans, grounded in empirical studies of human-robot interaction. By integrating cognitive hypotheses into AI design, Bennetot advances both the transparency of autonomous systems and their ability to collaborate effectively with people. His research is particularly valuable for students and practitioners interested in building AI that not only performs tasks but also explains its own learning journey.