Gabriel Bénédict

I am a PhD student conducting research at the intersection of theoretical and applied AI, working with the University of Amsterdam and RTL Nederland. My research focuses on several key areas:

• Metrics-as-losses for neural networks: Developing novel approaches to incorporate evaluation metrics directly into neural network training objectives
• Normative diversity metrics for news recommendation: Creating frameworks to measure and improve diversity in news recommendation systems
• Intent-satisfaction modeling: Building models that better understand and satisfy user intent in information retrieval
• Video-to-music AI: Working on the ProsAIc project to develop AI systems that can generate music from video content

My work combines theoretical insights with practical applications, particularly in the context of media and information systems.

• Artificial Intelligence: Neural networks, machine learning, deep learning
• Information Retrieval: Recommendation systems, search algorithms, user intent modeling
• Media Technology: News recommendation, content diversity, video-to-music AI
• Applied Statistics: Robust association measures, generalized linear models
• Economics: Development economics, debt relief, counterfactual analysis