Jordi Grau-Moya

I am a Research Scientist at Google DeepMind where I focus on understanding the fundamental mechanisms that allow the analysis and creation of safe AI agents. I use techniques from diverse areas such as reinforcement learning, causality, information theory and statistical mechanics.

  • Feel free to contact me: jordi dot grau dot mo at

  • List of publications: Google Scholar

Short Bio

In June 2020 I started working at DeepMind on AI Safety focusing on analysis and creation of safe AI agents. From 2017 to June 2020, I worked as a Senior Machine Learning Researcher at Secondmind (formerly, Cambridge (UK), developing new techniques that enable tuneable behaviour of artificial agents. In 2017 I completed my PhD thesis in Machine Learning and Computational Neuroscience at the Max Planck Institute for Intelligent Systems and at the Max Planck Institute for Biological Cybernetics in the beautiful Tübingen (Germany). Towards the end of my PhD I was an intern at Mitsubishi Electric Research Laboratories in Boston (USA) working on deep reinforcement learning. In 2011, I obtained my Engineering degree from Polytechnic University of Catalonia that concluded with my thesis about spiking neural network dynamics at the DONLL group at the Physics and Nuclear Engineering Department.