OpenAI works on advancing AI capabilities, safety, and policy.

Milestone Releases
Research Papers

May 5, 2020
Measuring the Algorithmic Efficiency of Neural Networks [Blog]

April 30, 2020
Jukebox: A Generative Model for Music [Blog]

April 16, 2020
Toward Trustworthy AI Development: Mechanisms for Supporting Verifiable Claims [Blog]

January 23, 2020
Scaling Laws for Neural Language Models

December 4, 2019
Deep Double Descent: Where Bigger Models and More Data Hurt [Blog]

December 3, 2019
Leveraging Procedural Generation to Benchmark Reinforcement Learning [Blog]

November 21, 2019
Benchmarking Safe Exploration in Deep Reinforcement Learning

November 13, 2019
Release Strategies and the Social Impacts of Language Models [Blog]

October 16, 2019
Solving Rubik's Cube with a Robot Hand [Blog]

September 18, 2019
Fine-Tuning Language Models from Human Preferences [Blog]

September 17, 2019
Emergent Tool Use From Multi-Agent Autocurricula [Blog]

August 21, 2019
Testing Robustness Against Unforeseen Adversaries [Blog]

July 10, 2019
The Role of Cooperation in Responsible AI Development [Blog]

May 28, 2019
SGD on Neural Networks Learns Functions of Increasing Complexity

May 3, 2019
Transfer of Adversarial Robustness Between Perturbation Types

April 23, 2019
Generating Long Sequences with Sparse Transformers [Blog]

March 20, 2019
Implicit Generation and Generalization in Energy-Based Models [Blog]

March 2, 2019
Neural MMO: A Massively Multiagent Game Environment for Training and Evaluating Intelligent Agents [Blog]
Reinforcement Learning

February 4, 2019
Computational Limitations in Robust Classification and Win-Win Results

December 14, 2018
An Empirical Model of Large-Batch Training [Blog]
Reinforcement Learning

December 6, 2018
Quantifying Generalization in Reinforcement Learning [Blog]
Reinforcement Learning

November 7, 2018
Concept Learning with Energy-Based Models [Blog]
Reinforcement Learning

November 5, 2018
Plan Online, Learn Offline: Efficient Learning and Exploration via Model-Based Control
Reinforcement Learning

October 30, 2018
Exploration by Random Network Distillation [Blog]
Reinforcement Learning

October 19, 2018
Supervising Strong Learners by Amplifying Weak Experts [Blog]
Reinforcement Learning

October 3, 2018
FFJORD: Free-Form Continuous Dynamics for Scalable Reversible Generative Models
Generative Models

October 1, 2018
Domain Randomization and Generative Models for Robotic Grasping

August 16, 2018
Constant Arboricity Spectral Sparsifiers

August 13, 2018
Large-Scale Study of Curiosity-Driven Learning
Reinforcement Learning

August 1, 2018
Learning Dexterous In-Hand Manipulation

July 31, 2018
Learning Policy Representations in Multiagent Systems
Reinforcement Learning

July 26, 2018
Variational Option Discovery Algorithms
Reinforcement Learning

July 9, 2018
Learning with Opponent-Learning Awareness
Reinforcment Learning

July 9, 2018
Glow: Generative Flow with Invertible 1x1 Convolutions [Blog]
Generative Models

June 2, 2018
GamePad: A Learning Environment for Theorem Proving

May 2, 2018
AI Safety via Debate [Blog]

April 25, 2018
Emergence of Grounded Compositional Language in Multi-Agent Populations
Reinforcement Learning

April 10, 2018
Gotta Learn Fast: A New Benchmark for Generalization in RL
Reinforcement Learning

April 4, 2018
On First-Order Meta-Learning Algorithms
Reinforcement Learning

March 19, 2018
Variance Reduction for Policy Gradient with Action-Dependent Factorized Baselines
Reinforcement Learning

March 14, 2018
Improving GANs Using Optimal Transport
Generative Models

March 8, 2018
Reptile: a Scalable Metalearning Algorithm
Reinforcement Learning

March 3, 2018
Sim-to-real Transfer of Robotic Control with Dynamics Randomization

March 3, 2018
Some Considerations on Learning to Explore via Meta-Reinforcement Learning
Reinforcement Learning

February 26, 2018
Multi-Goal Reinforcement Learning: Challenging Robotics Environments and Request for Research
Reinforcement Learning

February 23, 2018
Backpropagation through the Void: Optimizing Control Variates for Black-Box Gradient Estimation
Reinforcement Learning

February 20, 2018
The Malicious Use of Artificial Intelligence: Forecasting, Prevention, and Mitigation [Blog]

February 13, 2018
Evolved Policy Gradients [Blog]
Reinforcement Learning

February 2, 2018
DeepType: Multilingual Entity Linking by Neural Type System Evolution [Blog]
Reinforcement Learning

December 4, 2017
Learning Sparse Neural Networks through L0 Regularization
Reinforcement Learning

November 2, 2017
Interpretable and Pedagogical Examples

October 26, 2017
Meta Learning Shared Hierarchies [Blog]
Reinforcement Learning

October 17, 2017
Domain Randomization and Generative Models for Robotic Grasping

October 17, 2017
Asymmetric Actor Critic for Image-Based Robot Learning

October 12, 2017
Emergent Complexity via Multi-Agent Competition [Blog]
Reinforcement Learning

October 10, 2017
Continuous Adaptation via Meta-Learning in Nonstationary and Competitive Environments [Blog]
Reinforcement Learning

September 13, 2017
Learning with Opponent-Learning Awareness [Blog]

August 28, 2017
Proximal Policy Optimization Algorithms [Blog]
Reinforcement Learning

July 5, 2017
Hindsight Experience Replay
Reinforcement Learning

July 1, 2017
Teacher-Student Curriculum Learning
Reinforcement Learning

June 12, 2017
Deep reinforcement learning from human preferences [Blog]

June 7, 2017
Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments [Blog]

June 6, 2017
Parameter Space Noise for Exploration
Reinforcement Learning

June 5, 2017
UCB Exploration via Q-Ensembles
Reinforcement Learning

April 21, 2017
Equivalence Between Policy Gradients and Soft Q-Learning
Reinforcement Learning

April 10, 2017
Stochastic Neural Networks for Hierarchical Reinforcement Learning
Reinforcement Learning

April 5, 2017
Learning to Generate Reviews and Discovering Sentiment [Blog]

March 21, 2017
One-shot Imitation Learning

March 20, 2017
Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World [Blog]

March 15, 2017
Emergence of Grounded Compositional Language in Multi-Agent Populations [Blog]

March 12, 2017
Prediction and Control with Temporal Segment Models
Generative Models

March 10, 2017
Evolution Strategies as a Scalable Alternative to Reinforcement Learning [Blog]

March 6, 2017
Third Person Imitation Learning

February 8, 2017
Adversarial Attacks on Neural Network Policies

January 19, 2017
PixelCNN++: Improving the PixelCNN with Discretized Logistic Mixture Likelihood and Other Modifications
Generative Models

December 5, 2016
Reinforcement Learning

November 15, 2016
#Exploration: A Study of Count-Based Exploration for Deep Reinforcement Learning
Reinforcement Learning

November 14, 2016
On the Quantitative Analysis of Decoder-Based Generative Models
Generative Models

November 11, 2016
A Connection between Generative Adversarial Networks, Inverse Reinforcement Learning, and Energy-Based Models
Generative Models

November 9, 2016
RL2: Fast Reinforcement Learning via Slow Reinforcement Learning
Reinforcement Learning

November 8, 2016
Variational Lossy Autoencoder
Generative Models

November 7, 2016
Adversarial Training Methods for Semi-Supervised Text Classification

November 2, 2016
Extensions and Limitations of the Neural GPU

October 18, 2016
Semi-supervised Knowledge Transfer for Deep Learning from Private Training Data

October 11, 2016
Transfer from Simulation to Real World through Learning Deep Inverse Dynamics Model

August 29, 2016
Infrastructure for Deep Learning

June 21, 2016
Concrete Problems in AI Safety [Blog]

June 15, 2016
Improving Variational Inference with Inverse Autoregressive Flow [Blog]
Generative Models

June 12, 2016
InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets [Blog]
Generative Models

June 10, 2016
Improved Techniques for Training GANS [Blog]
Generative Models

June 5, 2016
OpenAI Gym [Blog]
Reinforcement Learning

June 4, 2016
Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks

May 31, 2016
VIME: Variational Information Maximizing Exploration [Blog]
Generative Models