Faster physics in Python
OpenAI is releasing an open-source Python library for simulating robots. The library uses the MuJoCo engine and was developed during their past year o...
OpenAI is releasing an open-source Python library for simulating robots. The library uses the MuJoCo engine and was developed during their past year o...
The article discusses the significance of the 2016 match between AlphaGo and professional Go player Lee Sedol. The game was considered intractable for...
Researchers at OpenAI have developed an algorithm that infers human preferences from simple feedback. By comparing two proposed behaviors and identify...
Researchers are exploring multiagent environments where agents compete for resources as a stepping stone towards developing Artificial General Intelli...
Researchers have proposed a new approach to exploration in reinforcement learning, called UCB (Upper Confidence Bound) exploration via Q-ensembles. Th...
OpenAI has released an open-source version of their Baselines project, which aims to reproduce state-of-the-art reinforcement learning algorithms. The...
Researchers have developed a robotic system that can learn a new task after observing it only once. The system is trained entirely in simulation and t...
OpenAI has released Roboschool, an open-source software tool for simulating robots. It integrates with the OpenAI Gym environment, which is a framewor...
Researchers have found an equivalence between two reinforcement learning algorithms: policy gradients and soft Q-learning. This means that the two met...
Researchers at OpenAI have developed a new approach to hierarchical reinforcement learning using stochastic neural networks. This method allows for mo...
OpenAI researchers have created a neural network that can learn to recognize sentiment in text without being explicitly trained for it. The model is t...
Researchers at OpenAI have developed an artificial intelligence system that can detect spam in the physical world. The AI was trained entirely in simu...