OpenAI

Source: https://openai.com

Found 1065 news

  • OpenAI Five Benchmark: Results
    OpenAI - 07:00 Aug 06, 2018
    Yesterday, OpenAI Five won a best-of-three against a team of 99.95th percentile Dota players: Blitz, Cap, Fogged, Merlini, and MoonMeander—four of whom have played Dota professionally—in front of a live audience and 100,000 concurrent livestream viewers.
  • Learning dexterity
    OpenAI - 07:00 Jul 30, 2018
    We’ve trained a human-like robot hand to manipulate physical objects with unprecedented dexterity.
  • Variational option discovery algorithms
    OpenAI - 07:00 Jul 26, 2018
  • OpenAI Scholars 2018: Meet our Scholars
    OpenAI - 07:00 Jul 25, 2018
    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.
    Tags: OpenAI Scholars
  • OpenAI Five Benchmark
    OpenAI - 07:00 Jul 18, 2018
    The OpenAI Five Benchmark match is now over!
  • Glow: Better reversible generative models
    OpenAI - 07:00 Jul 09, 2018
    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 generate realistic high resolution images, supports efficient sampling, and discovers features that can be used to manipulate attributes of data. We’re releasing code for the model and an online visualization tool so people can explore and build on these results.
    Tags: Glow
  • Learning Montezuma’s Revenge from a single demonstration
    OpenAI - 07:00 Jul 04, 2018
    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 plays a sequence of games starting from carefully chosen states from the demonstration, and learns from them by optimizing the game score using PPO, the same reinforcement learning algorithm that underpins OpenAI Five.
    Tags: Revenge
  • OpenAI Five
    OpenAI - 07:00 Jun 25, 2018
    Our team of five neural networks, OpenAI Five, has started to defeat amateur human teams at Dota 2.
    Tags: OpenAI Five
  • Retro Contest: Results
    OpenAI - 07:00 Jun 22, 2018
    The first run of our Retro Contest—exploring the development of algorithms that can generalize from previous experience—is now complete.
    Tags: Retro Contest
  • Learning policy representations in multiagent systems
    OpenAI - 07:00 Jun 17, 2018
  • Improving language understanding with unsupervised learning
    OpenAI - 07:00 Jun 11, 2018
    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: transformers and unsupervised pre-training. These results provide a convincing example that pairing supervised learning methods with unsupervised pre-training works very well; this is an idea that many have explored in the past, and we hope our result motivates further research into applying this idea on larger and more diverse datasets.
  • GamePad: A learning environment for theorem proving
    OpenAI - 07:00 Jun 02, 2018
  • OpenAI Fellows Fall 2018
    OpenAI - 07:00 May 30, 2018
    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.
  • Gym Retro
    OpenAI - 07:00 May 25, 2018
    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 over 1,000 games across a variety of backing emulators. We’re also releasing the tool we use to add new games to the platform.
    Tags: Gym Retro
  • AI and compute
    OpenAI - 07:00 May 16, 2018
    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’s Law had a 2-year doubling period)[^footnote-correction]. Since 2012, this metric has grown by more than 300,000x (a 2-year doubling period would yield only a 7x increase). Improvements in compute have been a key component of AI progress, so as long as this trend continues, it’s worth preparing for the implications of systems far outside today’s capabilities.
  • AI safety via debate
    OpenAI - 07:00 May 03, 2018
    We’re proposing an AI safety technique which trains agents to debate topics with one another, using a human to judge who wins.
  • Evolved Policy Gradients
    OpenAI - 07:00 Apr 18, 2018
    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. Agents trained with EPG can succeed at basic tasks at test time that were outside their training regime, like learning to navigate to an object on a different side of the room from where it was placed during training.
    Tags: Evolved Policy Gradients
  • Gotta Learn Fast: A new benchmark for generalization in RL
    OpenAI - 07:00 Apr 10, 2018
  • Retro Contest
    OpenAI - 07:00 Apr 05, 2018
    We’re launching a transfer learning contest that measures a reinforcement learning algorithm’s ability to generalize from previous experience.
  • Variance reduction for policy gradient with action-dependent factorized baselines
    OpenAI - 07:00 Mar 20, 2018

1065 items

Popular Tags