Articles Collectés

1,002

Total

941

Non traités

61

Traités

Non traitée 15/05/2017 09:00
Roboschool

RSS: OpenAI News

We are releasing Roboschool: open-source software for robot simulation, integrated with OpenAI Gym.

Non traitée 21/04/2017 09:00
Non traitée 10/04/2017 09:00
Non traitée 06/04/2017 09:00
Unsupervised sentiment neuron

RSS: OpenAI News

We’ve developed an unsupervised system which learns an excellent representation of sentiment, despite being trained only to predict the next character in the text of Amazon reviews.

Non traitée 01/04/2017 09:00
Spam detection in the physical world

RSS: OpenAI News

We’ve created the world’s first Spam-detecting AI trained entirely in simulation and deployed on a physical robot.

Non traitée 24/03/2017 08:00
Evolution strategies as a scalable alternative to reinforcement learning

RSS: OpenAI News

We’ve discovered that evolution strategies (ES), an optimization technique that’s been known for decades, rivals the performance of standard reinforcement learning (RL) techniques on modern RL ben...

Non traitée 21/03/2017 08:00
Non traitée 20/03/2017 08:00
Distill

RSS: OpenAI News

We’re excited to support today’s launch of Distill, a new kind of journal aimed at excellent communication of machine learning results (novel or existing).

Non traitée 16/03/2017 08:00
Learning to communicate

RSS: OpenAI News

In this post we’ll outline new OpenAI research in which agents develop their own language.

Non traitée 15/03/2017 08:00
Non traitée 12/03/2017 09:00
Non traitée 06/03/2017 09:00
Non traitée 24/02/2017 09:00
Attacking machine learning with adversarial examples

RSS: OpenAI News

Adversarial examples are inputs to machine learning models that an attacker has intentionally designed to cause the model to make a mistake; they’re like optical illusions for machines. In this post...

Non traitée 08/02/2017 09:00
Non traitée 30/01/2017 09:00
Team update

RSS: OpenAI News

The OpenAI team is now 45 people. Together, we’re pushing the frontier of AI capabilities—whether by validating novel ideas, creating new software systems, or deploying machine learning on robots.

Non traitée 19/01/2017 09:00
Non traitée 21/12/2016 09:00
Faulty reward functions in the wild

RSS: OpenAI News

Reinforcement learning algorithms can break in surprising, counterintuitive ways. In this post we’ll explore one failure mode, which is where you misspecify your reward function.

Non traitée 05/12/2016 09:00
Universe

RSS: OpenAI News

We’re releasing Universe, a software platform for measuring and training an AI’s general intelligence across the world’s supply of games, websites and other applications.

Non traitée 15/11/2016 09:00
Non traitée 15/11/2016 09:00
OpenAI and Microsoft

RSS: OpenAI News

We’re working with Microsoft to start running most of our large-scale experiments on Azure.