Manfred Diaz

Ph. D candidate at Mila and Montreal Robotics.

prof_pic.jpg

 Montreal, QC, Canada

I am a final year PhD candidate in Machine Learning and Robotics at Mila under the supervision of Liam Paull. My research focuses on AI and ML roots within other scientific disciplines, such as economics, game theory, mechanism design, and social choice theory, how various problems at the foundations of ML mirror others in these disciplines, and how these connections offer well-grounded frameworks for better understanding the present and shaping the future of AI.

In the past, I have interned at Google X, J.P. Morgan AI Research, Motional Inc., and Huawei Noah’s Ark Lab Canada.

Earlier, I graduated from Concordia University with an M.Sc. in Computer Science under Thomas Fevens while simultaneously acting as visiting researcher in the Shared Reality Lab at McGill University under the supervision of Jeremy Cooperstock. Before, I completed a B.Sc. in Computer Science summa cum laude from Universidad de las Ciencias Informaticas in Havana, Cuba and spent close to 10 years working as a software engineer and architect developing distributed systems with .NET Framework, Java and C++.

 Updates


May 22, 2025 Best paper award! Soft Condorcet Optimization won a best paper award at AAMAS 2025! See Marc’s announcement here.
May 19, 2025 Tutorial @AAMAS 2025 Marc Lanctot, Kate Larson and Ian Gemp presented the tutorial on Evaluation of General AI Agents at AAMAS 2025! We share the website, the slides and a Google Colab that we prepared for the occasion!
May 09, 2025 New Paper. Following our work on a theory of appropriateness for generative AI, in this new paper, we argue that AI safety and alignment should be focused on how humans deal with the irresoluble disagreements that persist in societies and the mechanisms used to prevent them from spiralling into conflict. It is also a blogpost on LessWrong!
May 06, 2025 Ph.D. Done! I successfully defended my Ph.D. thesis Machine Learning Through The Science of The Artificial at Mila and the University of Montreal. Many thanks to my advisor Liam Paull and to the committee: Michael Dennis, Pablo Samuel Castro and Gauthier Gidel for their great questions and feedback!
Feb 03, 2025 AAMAS 2025 Tutorial! Together with Marc Lanctot, Kate Larson and Ian Gemp we are presenting a tutorial on Evaluation of General AI Agents at AAMAS 2025! Website is here and tutorial notes are coming soon!

  Recent Publications


  1. A theory of appropriateness with applications to generative artificial intelligence
    Joel Z Leibo, Alexander Sasha Vezhnevets, Manfred Diaz, John P Agapiou, and 10 more authors
    arXiv [cs.AI], Dec 2024
  2. Rethinking Teacher-Student Curriculum Learning through the Cooperative Mechanics of Experience
    Manfred DiazLiam Paull, and Andrea Tacchetti
    Transactions on Machine Learning Research (TMLR), Dec 2024
  3. Soft Condorcet Optimization for Ranking of General Agents
    Marc Lanctot, Kate Larson, Michael Kaisers, Quentin Berthet, and 6 more authors
    arXiv [cs.MA], Oct 2024
  4. Milnor-Myerson Games and The Principles of Artificial Principal-Agent Problems
    Manfred DiazJoel Z Leibo, and Liam Paull
    In Finding The Frame: An RLC Workshop for Examining Conceptual Frameworks, Oct 2024
  5. Braxlines: Fast and Interactive Toolkit for RL-driven Behavior Engineering beyond Reward Maximization
    Shixiang Shane GuManfred Diaz, Daniel C. Freeman, Hiroki Furuta, and 6 more authors
    Oct 2021
  6. Active Domain Randomization
    Bhairav MehtaManfred Diaz, Florian Golemo, Christopher J. Pal, and 1 more author
    In Proceedings of the Conference on Robot Learning, Oct 2020