Meire Fortunato
Dr. Meire Fortunato is a Staff Research Scientist at Google DeepMind. She holds a Ph.D. in Mathematics from UC Berkeley, and her research focuses on modeling complex systems—spanning sequence models and reinforcement learning (Pointer Networks, Noisy Networks), geometric deep learning, and physical simulation (MeshGraphNets). She is a co-first author of GraphCast, an AI weather forecasting system featured on the cover of Science and awarded the 2024 MacRobert Award. Currently, her work centers on reasoning, scientific discovery, and agent systems. Dedicated to contributing to the research community at both a local and global scale, she is a co-founder of Khipu.ai, an organization strengthening the Latin American AI ecosystem, and has served in community leadership roles including Diversity & Inclusion Co-Chair for WiML at NeurIPS.
Research Areas
Authored Publications
Sort By
Noisy Networks for Exploration
Mohammad Gheshlaghi Azar
Bilal Piot
Jacob Menick
Ian Osband
Alexander Graves
Vlad Mnih
Remi Munos
Demis Hassabis
Olivier Pietquin
Charles Blundell
Shane Legg
Proceedings of the International Conference on Representation Learning (ICLR 2018), Vancouver (Canada)
Preview abstract
We introduce NoisyNet, a deep reinforcement learning agent with parametric noise added to its weights, and show that the induced stochasticity of the agent's policy can be used to aid efficient exploration.
The parameters of the noise are learned with gradient descent along with the remaining network weights. NoisyNet is straightforward to implement and adds little computational overhead.
We find that replacing the conventional exploration heuristics for A3C, DQN and dueling agents (entropy reward and epsilon-greedy respectively) with NoisyNet yields substantially higher scores for a wide range of Atari games, in some cases advancing the agent from sub to super-human performance.
View details
