Hi! I’m Sulin. CV

I recently joined MIT as a postdoctoral researcher working with Rafael Gómez-Bombarelli on developing machine learning methods for scientific settings. I received my PhD from Princeton University , advised by Ryan P. Adams and Peter J. Ramadge . My PhD research focuses on developing deep-learning-enabled probabilistic inference and generative modeling for knowledge discovery.

Previously, I worked with Sinno Jialin Pan on distrbuted (federated) multi-task learning at Nanyang Technological University in Singapore . Before that, I did my undergraduate in Electrical Engineering at National University of Singapore .

Research Areas

Generative marginal modeling
Approximating marginal probabilities of discrete data with neural networks, via a scalable training objective, for both maximum likelihood training and energy-based training (such as modeling Ising models, canonical ensemble of high-entropy alloys).

Deep-learning-enabled probabilistic inference
Scaling up probabilistic inference without sacrificing effectiveness, via using deep learning to accelerate the computationally expensive procedure in probabilistic inference.

Controllable discovery
Searching for interpretable/simple solutions or learning to control complex systems with safety guarantees.

Selected Publications

Generative Marginalization Models. Paper | Code | Video
Sulin Liu, Peter J. Ramadge, Ryan P. Adams.
Submitted, 2023.
short version at ICML Workshop on Workshop on Structured Probabilistic Inference & Generative Modeling. (Contributed talk, 6/125)

Sparse Bayesian Optimization. Paper | Code | Tutorial | Video
Sulin Liu* (equal contr.), Qing Feng*, David Eriksson*, Benjamin Letham, Eytan Bakshy.
International Conference on Artificial Intelligence and Statistics (AISTATS), 2023.
short version at at NeurIPS Workshop on Gaussian Processes, Spatiotemporal Modeling, and Decision-making Systems. (Contributed talk, top five selected)

Task-Agnostic Amortized Inference of Gaussian Process Hyperparameters. Paper | Code | Slides | Video
Sulin Liu, Xingyuan Sun, Peter J. Ramadge, Ryan P. Adams.
Advances in Neural Information Processing Systems (NeurIPS), 2020.
short version at 7th ICML Workshop on Automated Machine Learning. (Spotlight talk)

Site last updated on 2023-10-29