natasha jaques phd thesis defense

Citations per year

Duplicate citations, merged citations, add co-authors co-authors, cited by view all.

• Who’s Teaching What
• Subject Updates
• MEng program
• Opportunities
• Minor in Computer Science
• Resources for Current Students
• Program objectives and accreditation
• Graduate program requirements
• Admission process
• Degree programs
• Graduate research
• EECS Graduate Funding
• Resources for current students
• Student profiles
• Instructors
• DEI data and documents
• Recruitment and outreach
• Community and resources
• Get involved / self-education
• Rising Stars in EECS
• Graduate Application Assistance Program (GAAP)
• MIT Summer Research Program (MSRP)
• Sloan-MIT University Center for Exemplary Mentoring (UCEM)
• Electrical Engineering
• Computer Science
• Artificial Intelligence + Decision-making
• AI and Society
• AI for Healthcare and Life Sciences
• Artificial Intelligence and Machine Learning
• Biological and Medical Devices and Systems
• Communications Systems
• Computational Biology
• Computational Fabrication and Manufacturing
• Computer Architecture
• Educational Technology
• Electronic, Magnetic, Optical and Quantum Materials and Devices
• Graphics and Vision
• Human-Computer Interaction
• Information Science and Systems
• Integrated Circuits and Systems
• Nanoscale Materials, Devices, and Systems
• Natural Language and Speech Processing
• Optics + Photonics
• Optimization and Game Theory
• Programming Languages and Software Engineering
• Quantum Computing, Communication, and Sensing
• Security and Cryptography
• Signal Processing
• Systems and Networking
• Systems Theory, Control, and Autonomy
• Theory of Computation
• Departmental History
• Departmental Organization
• Visiting Committee
• News & Events
• News & Events
• EECS Celebrates Awards

Natasha Jaques – Social Reinforcement Learning

Grier A (34-401A)

Abstract: Social learning helps humans and animals rapidly adapt to new circumstances, coordinate with others, and drives the emergence of complex learned behaviors. What if it could do the same for AI? This talk describes how Social Reinforcement Learning in multi-agent and human-AI interactions can address fundamental issues in AI such as learning and generalization, while improving social abilities like coordination. I propose a unified method for improving coordination and communication based on causal social influence. I then demonstrate that multi-agent training can be a useful tool for improving learning and generalization. I present PAIRED, in which an adversary learns to construct training environments to maximize regret between a pair of learners, leading to the generation of a complex curriculum of environments. Agents trained with PAIRED generalize more than 20x better to unknown test environments. Finally, I demonstrate the value of learning socially from interacting with other agents, whether those agents are AI or humans. Together, this work argues that Social RL is a valuable approach for developing more general, sophisticated, and cooperative AI, which is ultimately better able to serve human needs.

Bio: Natasha Jaques holds a joint position as a Senior Research Scientist at Google Brain and Visiting Postdoctoral Scholar at UC Berkeley. Her research focuses on Social Reinforcement Learning in multi-agent and human-AI interactions. Natasha completed her PhD at MIT, where her thesis received the Outstanding PhD Dissertation Award from the Association for the Advancement of Affective Computing. Her work has also received Best Demo at NeurIPS, an honourable mention for Best Paper at ICML, Best of Collection in the IEEE Transactions on Affective Computing, and Best Paper at the NeurIPS workshops on ML for Healthcare and Cooperative AI. She has interned at DeepMind, Google Brain, and was an OpenAI Scholars mentor. Her work has been featured in Science Magazine, Quartz, IEEE Spectrum, MIT Technology Review, Boston Magazine, and on CBC radio. Natasha earned her Masters degree from the University of British Columbia, and undergraduate degrees in Computer Science and Psychology from the University of Regina.

• Date: Thursday, March 31
• Time: 2:00 pm - 3:00 pm
• Category: Special Seminar
• Location: Grier A (34-401A)
• Email: [email protected]

Natasha Jaques wins AAAC Outstanding PhD Dissertation Award 2021

Creative Commons

Attribution-NonCommercial-ShareAlike 4.0 International

Natasha Jaques

by Sarah Beckmann

Oct. 4, 2021

• Natasha Jaques Former Research Assistant

Share this post

Natasha Jaques, an alum of the Affective Computing group, has won this year's Association for the Advancement of Affective Computing (AAAC)  Outstanding PhD Dissertation Award for her thesis: " Social and Affective Machine Learning ." 

The AAAC Outstanding PhD Dissertation Award recognizes the most outstanding research contributions from recently graduated PhD students within the Affective Computing community. It was presented to Jaques at this year's  International Conference on Affective Computing and Intelligent Interaction (ACII)  on October 1, 2021. 

AI Songsmith Cranks Out Surprisingly Catchy Tunes

Google’s songwriting program learns by combining statistical learning and explicit rules—the same approach may make it easier for engineers…

Natasha Jaques Dissertation Defense

Towards Social and Affective Artificial IntelligenceSocial learning is a crucial component of human intelligence, allowing us to rapidly ad…

Tuning Recurrent Neural Networks with Reinforcement Learning

Jaquesn, N., Gu, S., Turner, R., and Eck, D. International Conference on Learning Representations (ICLR) workshop, Toulon, France, April 2017

Sequence Tutor: Conservative Fine-Tuning of Sequence Generation Models with KL-Control

Jaques, N., Gu, S., Bahdanau, D., Hernandez-Lobato, J., Turner, R., and Eck, D. International Conference on Machine Learning (ICML), Sydney, Australia, August 2017.

Natasha Jaques

Assistant Professor, Paul G. Allen School of Computer Science & Engineering

Research focus

Deep reinforcement learning, multi-agent reinforcement learning, reinforcement learning from human feedback, human-AI interaction, social learning

Ph.D. Media Arts and Sciences, Massachusetts Institute of Technology, 2019 M.Sc. Computer Science, University of British Columbia, 2014 B.Sc. Computer Science, University of Regina, 2012 B.A. Psychology, University of Regina, 2012

Natasha Jaques will join the Allen School this winter from Google Brain, where she is a senior research scientist exploring if AI agents benefit from social learning. She has also interned at DeepMind and Google Brain, and was an OpenAI Scholars mentor.

Jaques’s research focuses on social reinforcement learning in multi-agent and human-AI interactions. Her interest in AI learning and collaboration extends into developing multi-training algorithms that create automatic curriculum to help AI learn from each other, and improving mechanisms that allow AI to learn from human partners. Jaques has received numerous awards including Best Demo at NeurIPS, Best of Collection in the IEEE Transactions on Affective Computing, and Best Paper at the NeurIPS workshops on ML for Healthcare and Cooperate AI. Her work has been featured in Science Magazine, MIT Technology Review, Quartz, IEEE Spectrum, Boston Magazine and on CBC Radio.

Natasha Jaques

Research areas.

Human-Computer Interaction and Visualization

Machine Intelligence

Natural Language Processing

Research Area

• Health & Bioscience 1
• Human-Computer Interaction and Visualization 2
• Machine Intelligence 10
• Machine Perception 1
• Natural Language Processing 2
• Title, descending
• Year, descending

We're always looking for more talented, passionate people.

TalkRL: The Reinforcement Learning Podcast

Natasha jaques, featured references .

• MIT Media Lab Flight Offsets ,  Caroline Jaffe, Juliana Cherston, Natasha Jaques 
• Modeling Others using Oneself in Multi-Agent Reinforcement Learning ,  Roberta Raileanu, Emily Denton, Arthur Szlam, Rob Fergus 
• Inequity aversion improves cooperation in intertemporal social dilemmas ,   Edward Hughes, Joel Z. Leibo, Matthew G. Phillips, Karl Tuyls, Edgar A. Duéñez-Guzmán, Antonio García Castañeda, Iain Dunning, Tina Zhu, Kevin R. McKee, Raphael Koster, Heather Roff, Thore Graepel 
• Sequential Social Dilemma Games on github , Eugene Vinitsky, Natasha Jaques  
• AI Alignment newsletter , Rohin Shah 
• Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly Complex and Diverse Learning Environments and Their Solutions , Rui Wang, Joel Lehman, Jeff Clune, Kenneth O. Stanley 
• The social function of intellect , Nicholas Humphrey 
• Autocurricula and the Emergence of Innovation from Social Interaction: A Manifesto for Multi-Agent Intelligence Research , Joel Z. Leibo, Edward Hughes, Marc Lanctot, Thore Graepel 
• A Recipe for Training Neural Networks , Andrej Karpathy 
• Emotionally Adaptive Intelligent Tutoring Systems using POMDPs , Natasha Jaques 
• Sapiens , Yuval Noah Harari 

Creators and Guests

headphones Listen Anywhere

Natasha Jaques

Contact information:.

• [email protected]
• Affective Computing

PhD candidate working on  improving deep learning and AI agents by building in forms of affective and social intelligence. My past work has investigated methods for improving generalization of machine learning models via intrinsic motivation, transfer learning, multi-task learning, and learning from human preferences.  I've interned with DeepMind and Google Brain, and was an OpenAI Scholars mentor.  Experienced in traditional machine learning,  deep learning, kernel methods, Bayesian non-parametrics, causal inference, and reinforcement learning.

My favourite past projects have included: 

- Developing a unified method for promoting cooperation and communication among in multi-agent reinforcement learning (RL) by creating an intrinsic reward based on assessing causal influence between agents.

- Improving deep generative models by using human facial expression responses to samples from the model as a training signal.

 - Effectively combining supervised learning and RL to train generative sequence models.

- Using multi-task learning techniques to personalize machine learning models and improve accuracy in predicting next day stress, happiness and health.

Multi-task Learning for Predicting Health, Stress, and Happiness

Jaques, N., Taylor, S., Nosakhare, E., Sano, A., Picard, R. In Proc. NIPS Workshop on ML in Health, Barcelona, Spain, December 2016. **BEST PAPER AWARD**

Multimodal Autoencoder: A Deep Learning Approach to Filling In Missing Sensor Data and Enabling Better Mood Prediction

Jaques, N., Taylor, S., Sano, A., and Picard, R. International Conference on Affective Computing and Intelligent Interaction (ACII), San Antonio, Texas, October 2017

EDA Explorer

Electrodermal Activity (EDA) is a physiological indicator of stress and strong emotion. While an increasing number of wearable devices can …

Predicting Tomorrow's Mood, Health, and Stress Level using Personalized Multitask Learning and Domain Adaptation

Jaques, N., Rudovic, O., Taylor, S., Sano, A., and Picard, R. Proceedings of Machine Learning Research, 48, 17-33. August 2017.

Improving RNN Sequence Generation with RL

This project investigates a general method for improving the structure and quality of sequences generated by a recurrent neural network (RN…

Causal Influence Intrinsic Social Motivation for Multi-Agent Reinforcement Learning

Teaching multiple AI agents to coordinate their behavior represents a challenging task, that can be difficult to achieve without training a…

Personality, Attitudes, and Bonding in Conversations

Jaques, N., Kim, Y. L., and Picard, R. W. "Personality, Attitudes, and Bonding in Conversations," In Proceedings of Intelligent Virtual Agents, California, USA, September 2016.

Understanding and Predicting Bonding in Conversations Using Thin Slices of Facial Expressions and Body Language

Jaques, N., McDuff, D., Kim, Y. K., and Picard, R. W. "Understanding and Predicting Bonding in Conversations Using Thin Slices of Facial Expressions and Body Language," In Proceedings of Intelligent Virtual Agents, California, USA, September 2016.