Publications
2024
Scaling Instructable Agents Across Many Simulated Worlds
The SIMA Team
Tech Report, 2024
Tech Report Blog
2023
Bridging the Gap Between Target Networks and Functional Regularization
Alexandre Piché, Valentin Thomas, Rafael Pardinas, Joseph Marino, Gian Maria Marconi, Chris J. Pal, Mohammad Emtiyaz Khan
Transactions on Machine Learning Research (TMLR), Sept. 2023
arXiv
Insights from Generative Modeling for Neural Video Compression
Ruihan Yang, Yibo Yang, Joseph Marino, Stephan Mandt
Transactions on Pattern Analysis and Machine Intelligence (TPAMI), Apr. 2023
arXiv
2022
Predictive Coding, Variational Autoencoders, and Biological Connections
Joseph Marino
Neural Computation, Jan. 2022
arXiv publisher
2021
Iterative Amortized Policy Optimization
Joseph Marino, Alexandre Piché, Alessandro Davide Ialongo, Yisong Yue
NeurIPS, 2021
arXiv code
Improving Sequential Latent Variable Models with Autoregressive Flows
Joseph Marino, Lei Chen, Jiawei He, Stephan Mandt
Machine Learning, Nov. 2021
arXiv publisher code
Learned Feedback & Feedforward Perception & Control
Joseph Marino
Ph.D. Thesis, 2021
PDF
Beyond Target Networks: Improving Deep Q-learning with Functional Regularization
Alexandre Piché, Joseph Marino, Gian Maria Marconi, Chris J. Pal, Mohammad Emtiyaz Khan
Preprint, 2021
arXiv
Scale Space Flow with Autoregressive Priors
Ruihan Yang, Yibo Yang, Joseph Marino, Stephan Mandt
ICLR 2021 Workshop on Neural Compression
PDF
Hierarchical Autoregressive Modeling for Neural Video Compression
Ruihan Yang, Yibo Yang, Joseph Marino, Stephan Mandt
ICLR 2021
arXiv code
2020
Sequential Autoregressive Flow-Based Policies
Alex Guerra, Joseph Marino
ICML 2020 Workshop on Invertible Neural Networks, Normalizing Flows, and Explicit Likelihood Models
PDF code video
Deep Generative Video Compression with Temporal Autoregressive Transform
Ruihan Yang, Yibo Yang, Joseph Marino, Yang Yang, Stephan Mandt
ICML 2020 Workshop on Invertible Neural Networks, Normalizing Flows, and Explicit Likelihood Models
PDF
2019
Predictive Coding, Variational Autoencoders, and Biological Connections (Workshop Version)
Joseph Marino
NeurIPS 2019 Workshop on Real Neurons and Hidden Units, NAISys
arXiv PDF NeurIPS poster NAISys poster
Improving Sequential Latent Variable Models with Autoregressive Flows (Workshop Version)
Joseph Marino, Lei Chen, Jiawei He, Stephan Mandt
Advances in Approximate Bayesian Inference (AABI) 2019
arXiv PDF poster slides code
On the Design of Variational RL Algorithms
Joseph Marino, Alexandre Piché, Yisong Yue
NeurIPS 2019 Workshop on Deep Reinforcement Learning
PDF poster
An Inference Perspective on Model-Based Reinforcement Learning
Joseph Marino, Yisong Yue
ICML 2019 Workshop on Generative Modeling and Model-Based Reasoning for Robotics and AI
PDF poster slides talk @ 41m 30s
Variational Autoencoders with Jointly Optimized Latent Dependency Structure
Jiawei He, Yu Gong, Joseph Marino, Greg Mori, Andreas Lehrmann
ICLR 2019
PDF
2018
A General Method for Amortizing Variational Filtering
Joseph Marino, Milan Cvitkovic, Yisong Yue
NeurIPS 2018
arXiv PDF poster code
Variational Latent Dependency Learning
Jiawei He, Yu Gong, Joseph Marino, Greg Mori, Andreas Lehrmann
NeurIPS 2018 Workshop on Bayesian Deep Learning
PDF
Probabilistic Video Generation using Holistic Attribute Control
Jiawei He, Andreas Lehrmann, Joseph Marino, Greg Mori, Leonid Sigal
ECCV 2018
arXiv
Iterative Amortized Inference
Joseph Marino, Yisong Yue, Stephan Mandt
ICML 2018
arXiv PDF poster code video
A General Framework for Amortizing Variational Filtering
Joseph Marino, Milan Cvitkovic, Yisong Yue
ICML 2018 Workshop on Theoretical Foundations and Applications of Deep Generative Models
PDF poster
Learning to Infer
Joseph Marino, Yisong Yue, Stephan Mandt
ICLR 2018 Workshop Track
PDF poster
2017
Iterative Inference Models
Joseph Marino, Yisong Yue, Stephan Mandt
NeurIPS 2017 Workshop on Bayesian Deep Learning
PDF poster
2016
Non-expert Labels Improve Fine-Grained Object Recognition
Joseph Marino
Preprint 2016
PDF
Taxonomic Curriculum Learning for Fine-Grained Object Recognition
Joseph Marino
Preprint 2016
PDF