Articles Collectés

1,002

Total

941

Non traités

61

Traités

Non traitée 25/07/2018 09:00
OpenAI Scholars 2018: Meet our Scholars

RSS: OpenAI News

Our first class of OpenAI Scholars is underway, and you can now follow along as this group of experienced software developers becomes machine learning practitioners.

Non traitée 18/07/2018 09:00
OpenAI Five Benchmark

RSS: OpenAI News

The OpenAI Five Benchmark match is now over!

Non traitée 09/07/2018 09:00
Glow: Better reversible generative models

RSS: OpenAI News

We introduce Glow, a reversible generative model which uses invertible 1x1 convolutions. It extends previous work on reversible generative models and simplifies the architecture. Our model can gen...

Non traitée 04/07/2018 09:00
Learning Montezuma’s Revenge from a single demonstration

RSS: OpenAI News

We’ve trained an agent to achieve a high score of 74,500 on Montezuma’s Revenge from a single human demonstration, better than any previously published result. Our algorithm is simple: the agent...

Non traitée 25/06/2018 09:00
OpenAI Five

RSS: OpenAI News

Our team of five neural networks, OpenAI Five, has started to defeat amateur human teams at Dota 2.

Non traitée 22/06/2018 09:00
Retro Contest: Results

RSS: OpenAI News

The first run of our Retro Contest—exploring the development of algorithms that can generalize from previous experience—is now complete.

Non traitée 17/06/2018 09:00
Non traitée 11/06/2018 09:00
Improving language understanding with unsupervised learning

RSS: OpenAI News

We’ve obtained state-of-the-art results on a suite of diverse language tasks with a scalable, task-agnostic system, which we’re also releasing. Our approach is a combination of two existing ideas:...

Non traitée 02/06/2018 09:00
Non traitée 30/05/2018 09:00
OpenAI Fellows Fall 2018

RSS: OpenAI News

We’re now accepting applications for the next cohort of OpenAI Fellows, a program which offers a compensated 6-month apprenticeship in AI research at OpenAI.

Non traitée 25/05/2018 09:00
Gym Retro

RSS: OpenAI News

We’re releasing the full version of Gym Retro, a platform for reinforcement learning research on games. This brings our publicly-released game count from around 70 Atari games and 30 Sega games to ...

Non traitée 16/05/2018 09:00
AI and compute

RSS: OpenAI News

We’re releasing an analysis showing that since 2012, the amount of compute used in the largest AI training runs has been increasing exponentially with a 3.4-month doubling time (by comparison, Moore...

Non traitée 03/05/2018 09:00
AI safety via debate

RSS: OpenAI News

We’re proposing an AI safety technique which trains agents to debate topics with one another, using a human to judge who wins.

Non traitée 18/04/2018 09:00
Evolved Policy Gradients

RSS: OpenAI News

We’re releasing an experimental metalearning approach called Evolved Policy Gradients, a method that evolves the loss function of learning agents, which can enable fast training on novel tasks. Agen...

Non traitée 10/04/2018 09:00
Non traitée 05/04/2018 09:00
Retro Contest

RSS: OpenAI News

We’re launching a transfer learning contest that measures a reinforcement learning algorithm’s ability to generalize from previous experience.

Non traitée 20/03/2018 08:00
Non traitée 15/03/2018 08:00
Non traitée 15/03/2018 08:00
Report from the OpenAI hackathon

RSS: OpenAI News

On March 3rd, we hosted our first hackathon with 100 members of the artificial intelligence community.

Non traitée 08/03/2018 09:00