Ondrej Biza's repositories
action_priors
Source code for Action Priors for Large Action Spaces in Robotics. Biza, Wang, Platt, van de Meent, Wong. AAMAS'21.
action_attention_iclr_22
Source code for "Binding Actions to Objects in World Models. Ondrej Biza, Robert Platt, Jan-Willem van de Meent, Lawson L. S. Wong and Thomas Kipf. ICLR 2022 workshop on Objects, Structure and Causality."
negative-sampling-icml-21
Original source code for "The Impact of Negative Sampling on Contrastive Structured World Models. O Biza, E van der Pol, T Kipf. ICML'21 SSL workshop."
discrete_abstractions
Source code for Learning Discrete State Abstractions With Deep Variational Inference. Biza, Platt, van de Meent, Wong. AABI'21. https://arxiv.org/abs/2003.04300
baby-a3c
A high-performance Atari A3C agent in 180 lines of PyTorch
google-research
Google Research
ondrejbiza.github.io
My personal website.
pgm
Exercises from the Probabilistic Graphical Models: Principles and Techniques book.
pytorch-maml-rl
Reinforcement Learning with Model-Agnostic Meta-Learning in Pytorch
se3-transformer-public
code for the SE3 Transformers paper: https://arxiv.org/abs/2006.10503
sonnet
TensorFlow-based neural network library
SPACE
Official PyTorch implementation of "SPACE: Unsupervised Object-Oriented Scene Representation via Spatial Attention and Decomposition"
stable-baselines
A fork of OpenAI Baselines, implementations of reinforcement learning algorithms
taxpose
TAX-Pose: Task-Specific Cross-Pose Estimation for Robot Manipulation