I'm a PhD student at the MIT Media Lab, working with Deb Roy in the Laboratory for Social Machines group. I also work with Jacob Andreas in the CSAIL Language & Intelligence group.
My broad goal is to develop machine learning methods to improve communication (between humans, between humans and machines, and between machines). My current focus is on human-AI collaboration, and how model explanations can be best utilized by humans in algorithm-in-the-loop settings.
My earlier research includes text generative models, as applied to summarization, style transfer, text simplification, and dialogue generation. I’m also interested in unsupervised methods and methods for adapting to new data (continual learning, meta learning, transfer learning). Sometimes my research is done through the lens of stories. For my Master’s thesis, I worked on audio-visual sentiment analysis to learn emotional arcs in movies.
Previously, I've worked with Jason Weston and Stephen Roller at Facebook AI Research, and Peter J. Liu at Google Brain. Before that, I spent one year as a data scientist at Facebook, after graduating from UC Berkeley.
Projects (Selected) ¬
Natural Language Processing
Unsupervised Neural Multi-document Abstractive Summarization
Abstractive summarization of reviews without labeled data.
Towards Content Consistent Style Transfer with Disentangled Factors
Disentangling factors of variation in order to alter the sentiment and source attributes (speaker political orientation) of Yelp reviews and tweets.
Inferring personas from dialogue
Neural attentive memory networks for inferring personas from dialogues.
Fairness, Accountability, Transparency
Games for Fairness and Interpretability
Framework and examples of machine learning-powered Games with a Purpose to increase fairness and transparency while providing researchers with additional training data and adversarial examples.
AI & Equality Course
Co-organized seminar course on ethical implications of AI -- specifically how AI can both promote and impede equality in various domains (e.g. healthcare, law enforcement, labor markets, etc).
Stories, Art, & AI
Learning emotional arcs in movies
Audio-visual sentiment analysis for learning emotional arcs in movies and predicting audience engagement.
Artistic influencer GAN
Modeling and conditioning on an artistic influence graph to generate paintings.
Tool for automatically juxtaposing rap lyrics over classical paintings based on semantic and stylistic content.
Games for Fairness and Interpretability.
Eric Chu*, Nabeel Gillani*, and Sneha Priscilla Makini. The Web Conference 2020.
MeanSum: A Neural Model for Unsupervised Multi-Document Abstractive Summarization
Eric Chu* and Peter J. Liu*. ICML 2019.
Learning Personas from Dialogue with Attentive Memory Networks
Eric Chu*, Prashanth Vijayaraghavan*, and Deb Roy. Empirical Methods in Natural Language Processing, 2018.
Artistic Influence GAN
Eric Chu. Neural Information Processing Systems, Creativity Workshop 2018.
Audio-visual Sentiment Analysis for Learning Emotional Arcs in Movies
Eric Chu and Deb Roy. Data Mining (ICDM), 2017 IEEE 17th International Conference on. IEEE, 2017.
Human Atlas: A Tool for Mapping Social Networks.
Saveski, Martin, Eric Chu, Soroush Vosoughi, and Deb Roy. In Proceedings of the 25th International Conference Companion on World Wide Web, pp. 247-250. International World Wide Web Conferences Steering Committee, 2016.
Facebook AI Research (Summer 2019)
Google Brain (Summer 2018)
MIT Media Lab (2017 – now)
MIT Media Lab (2015 – 2017)
Data Scientist, Ads Integrity Team
UC Berkeley (2010-2014)
B.Sc Student, Electrical Engineering & Computer Science