Natasha Jaques

Natasha Jaques is an Assistant Professor at the University of Washington Paul G. Allen School of Computer Science & Engineering and a Senior Research Scientist at Google DeepMind. She leads the Social RL Lab, focusing on algorithms that integrate social learning with multi-agent reinforcement learning (RL) to enhance AI agents' learning, generalization, coordination, and human-AI interaction.

• PhD from MIT Media Lab (2019)
• MSc in Computer Science from University of British Columbia (2014)
• BSc/BA in Computer Science and Psychology from University of Regina (2012)

Her research spans three core areas:

• Social Reinforcement Learning - Developing frameworks like PAIRED for emergent complexity and zero-shot transfer through adversarial environment generation, and Social Influence for intrinsic motivation in multi-agent coordination.
• Human-AI Interaction - Innovating offline RL techniques for dialog systems using implicit signals like sentiment, and affective computing approaches for physiological data analysis.
• Machine Learning for Wellbeing - Pioneering personalized multi-task models to predict mood, stress, and health outcomes using hierarchical Bayes and domain adaptation.

Her scientific awards include:

• Best Paper, AAAI Representation Learning for Responsible Human-Centric AI workshop (2023)
• Outstanding PhD Dissertation, Association for the Advancement of Affective Computing (2021)
• Best of Collection, IEEE Transactions on Affective Computing (2021)
• Best Paper, NeurIPS Workshop on Cooperative AI (2020)
• Best Paper Honourable Mention, ICML (2019)
• Rising Stars in EECS Pitch Competition Winner (2019)

Jaques has made significant contributions to open-source tools like the EDA Explorer for wearable data analysis and PAIRED environment generation. Her work appears in top venues including NeurIPS, ICML, EMNLP, and CVPR, with press coverage in Science Magazine, MIT Technology Review, and CBC Radio.