I am about to become a third year Ph.D. student in Stanford University. I think about problems in machine learning and deep learning under the supervision of Stefano Ermon. My publications cover topics such as deep generative models, Bayesian inference and (inverse) reinforcement learning and their applications.

I did my undergrad at Tsinghua University, where I was lucky enough to collaborate with Jun Zhu and Lawrence Carin on scalable Bayesian machine learning.


Contact: tsong at cs.stanford.edu

Publications

A Lagrangian Perspective on Latent Variable Generative Models

Shengjia Zhao, Jiaming Song, Stefano Ermon
To appear in the 2018 Conference on Uncertainty in Artificial Intelligence (UAI 2018). [code]
Abridged version in ICML 2018 Workshop on Theoretical Foundations and Applications of Deep Generative Models


Accelerating Natural Gradient with Higher-Order Invariance

Yang Song, Jiaming Song, Stefano Ermon
To appear in the 2018 International Conference on Machine Learning (ICML 2018). [code]


Adversarial Constraint Learning for Structured Prediction

Hongyu Ren, Russell Stweart, Jiaming Song, Volodymyr Kuleshov, Stefano Ermon
To appear in the 2018 International Joint Conference on Artificial Intelligence (IJCAI 2018). [code]
Abridged version in NIPS 2017 Workshop on Learning with Limited Data


A-NICE-MC: Adversarial Training for MCMC

Jiaming Song, Shengjia Zhao, Stefano Ermon
In the 30th Neural Information Processing Systems (NIPS 2017). [code] [poster] [slides]
Abridged version in ICML 2017 Workshop on Implicit Models [one pager]


InfoGAIL: Interpretable Imitation Learning from Visual Demonstrations

Yunzhu Li, Jiaming Song, Stefano Ermon
In the 30th Neural Information Processing Systems conference (NIPS 2017). [code]


Learning Hierarchical Features from Generative Models

Shengjia Zhao, Jiaming Song, Stefano Ermon
In the 34th International Conference on Machine Learning (ICML 2017). [code] [blog]


Factored Sigmoid Belief Networks for Sequence Learning

Jiaming Song, Zhe Gan, Lawrence Carin
In the 33rd International Conference on Machine Learning (ICML 2016).


Organizational Churn: A Roll of the Dice?

Canyao Liu*, Jiaming Song*, Chuan Yu*.
Undergraduate Mathematics and its Applications (UMAP), Issue 36.2.


Discriminative Nonparametric Latent Feature Relational Models with Data Augmentation

Bei Chen, Ning Chen, Jun Zhu, Jiaming Song, Bo Zhang
In the 30th Association for the Advancement of Artificial Intelligence Conference (AAAI 2016).

Workshop Papers

Markov Chain Monte Carlo for Learning Belief Networks

Laëtitia Shao, Jiaming Song, Aditya Grover, Stefano Ermon
In ICML 2018 Workshop on Theoretical Foundations and Applications of Deep Generative Models


Learning Controllable Fair Representations via Latent Variable Generative Models

Jiaming Song, Pratyusha Kalluri, Aditya Grover, Shengjia Zhao and Stefano Ermon
In ICML 2018 Workshop on Theoretical Foundations and Applications of Deep Generative Models


Multi-Agent Generative Adversarial Imitation Learning

Jiaming Song, Hongyu Ren, Dorsa Sadigh, Stefano Ermon
In the 6th International Conference on Learning Representations (ICLR 2018) Workshop Track.


Structured Prediction with Adversarial Constraint Learning

Hongyu Ren, Russell Stweart, Jiaming Song, Volodymyr Kuleshov, Stefano Ermon
In NIPS 2017 Workshop on Learning with Limited Data


A Lagrangian Perspective on Latent Variable Generative Modeling

Shengjia Zhao, Jiaming Song, Stefano Ermon
In NIPS 2017 Workshop on Bayesian Deep Learning


An Empirical Study of the Generalization Behavior of Generative Adversarial Networks

Hongyu Ren, Shengjia Zhao, Jiaming Song, Lijie Fan, Stefano Ermon
In NIPS 2017 Workshop on Deep Learning: Bridging Theory and Practice


Generative Adversarial Learning of Markov Chains

Jiaming Song, Shengjia Zhao, Stefano Ermon
In the 5th International Conference on Learning Representations (ICLR 2017) Workshop Track. [code] [poster]

Preprints

An Empirical Analysis of Proximal Policy Optimization with Kronecker-factored Natural Gradients

Jiaming Song, Yuhuai Wu
arXiv preprint arXiv:1801.05566.


InfoVAE: Information Maximizing Variational Autoencoders

Shengjia Zhao, Jiaming Song, Stefano Ermon
arXiv preprint arXiv:1706.02262. [code]


Towards Deeper Understanding of Variational Autoencoding Models

Shengjia Zhao, Jiaming Song, Stefano Ermon
arXiv preprint arXiv:1702.08658. [code] [blog]


On the Limits of Learning Representations with Label-Based Supervision

Jiaming Song, Russell Stewart, Shengjia Zhao, Stefano Ermon
arXiv preprint arXiv:1703.02156


Max-Margin Nonparametric Latent Feature Relational Models for Link Prediction

Jun Zhu, Jiaming Song, Bei Chen
arXiv preprint arXiv:1602.07428
Under review of Artificial Intelligence Journal