Jiaming Song

InfoVAE: Information Maximizing Variational Autoencoders

Shengjia Zhao, Jiaming Song, Stefano Ermon
ArXiv
2017

A key advance in learning generative models is the use of amortized inference distributions that are jointly trained with the models. We find that existing training objectives for variational… (More)

InfoGAIL: Interpretable Imitation Learning from Visual Demonstrations

Algorithm 1 InfoGAIL Input: Initial parameters of policy, discriminator and posterior approximation θ0, ω0, ψ0; expert trajectories τE ∼ πE containing state-action pairs. Output: Learned policy πθ… (More)

Inferring The Latent Structure of Human Decision-Making from Raw Visual Inputs

The goal of imitation learning is to match example expert behavior, without access to a reinforcement signal. Expert demonstrations provided by humans, however, often show significant variability due… (More)

Towards Deeper Understanding of Variational Autoencoding Models

Shengjia Zhao, Jiaming Song, Stefano Ermon
ArXiv
2017

We propose a new family of optimization criteria for variational auto-encoding models, generalizing the standard evidence lower bound. We provide conditions under which they recover the data… (More)

A-NICE-MC: Adversarial Training for MCMC

Jiaming Song, Shengjia Zhao, Stefano Ermon
NIPS
2017

Existing Markov Chain Monte Carlo (MCMC) methods are either based on generalpurpose and domain-agnostic schemes, which can lead to slow convergence, or problem-specific proposals hand-crafted by an… (More)

Learning Hierarchical Features from Deep Generative Models

Shengjia Zhao, Jiaming Song, Stefano Ermon
ICML
2017

Deep neural networks have been shown to be very successful at learning feature hierarchies in supervised learning tasks. Generative models, on the other hand, have benefited less from hierarchical… (More)

Multi-Agent Generative Adversarial Imitation Learning

Jiaming Song, Hongyu Ren, Dorsa Sadigh, Stefano Ermon
NeurIPS
2018

Imitation learning algorithms can be used to learn a policy from expert demonstrations without access to a reward signal. However, most existing approaches are not applicable in multi-agent settings… (More)

A Lagrangian Perspective on Latent Variable Generative Models

Shengjia Zhao, Jiaming Song, Stefano Ermon
UAI
2018

A large number of objectives have been proposed to train latent variable generative models. We show that many of them are Lagrangian dual functions of the same primal optimization problem. The primal… (More)

Learning Hierarchical Features from Generative Models

Shengjia Zhao, Jiaming Song, Stefano Ermon
ArXiv
2017

Deep neural networks have been shown to be very successful at learning feature hierarchies in supervised learning tasks. Generative models, on the other hand, have benefited less from hierarchical… (More)

Discriminative Nonparametric Latent Feature Relational Models with Data Augmentation

Bei Chen, Ning Chen, Jun Zhu, Jiaming Song, Bo Zhang
AAAI
2016

We present a discriminative nonparametric latent feature relational model (LFRM) for link prediction to automatically infer the dimensionality of latent features. Under the generic RegBayes… (More)

Filter Results:

Publication Year

Publication Type

Co-author

Journals and Conferences