Joel Veness

Research Scientist, Google DeepMind

Journal Articles

  • Human-level control through deep reinforcement learning
    Volodymyr Mnih, Koray Kavukcuoglu, David Silver, Andrei A. Rusu, Joel Veness, Marc G. Bellemare, Alex Graves, Martin Riedmiller, Andreas K. Fidjeland, Georg Ostrovski, Stig Petersen, Charles Beattie, Amir Sadik, Ioannis Antonoglou, Helen King, Dharshan Kumaran, Daan Wierstra, Shane Legg, Demis Hassabis
    Nature, 2015
    official page

  • The Arcade Learning Environment:
        An Evaluation Platform for General Agents

    Marc Bellemare, Yavar Naddaf, Joel Veness, Michael Bowling
    Journal of Artificial Intelligence Research (JAIR), 2013
    pdf official page arXive ALE homepage bibtex

  • A Monte-Carlo AIXI Approximation
    Joel Veness, Kee Siong Ng, Marcus Hutter, William Uther, David Silver
    Journal of Artificial Intelligence Research (JAIR), 2011
    pdf official page pacman video youtube video source bibtex
    Honorable Mention for 2014 IJCAI-JAIR Best Paper Prize

Conference Papers

  • The Forget-me-not Process
    Neural Information Processing Systems (NIPS), 2016
    Joel Veness, Kieran Milan, James Kirkpatrick, Anna Koop, Michael Bowling, Demis Hassabis
    pdf appendix source code

  • Online Learning of k-CNF Boolean Functions
    Joel Veness, Marcus Hutter, Laurent Orseau, Marc Bellemare
    International Joint Conference on Artificial Intelligence (IJCAI), 2015
    pdf tech report bibtex

  • Compress and Control
    Joel Veness, Marc Bellemare, Marcus Hutter, Alvin Chua, Guillaume Desjardins
    Association for the Advancement of Artificial Intelligence (AAAI), 2015
    pdf bibtex

  • Skip Context Tree Switching
    Marc Bellemare, Joel Veness, Erik Talvitie
    International Conference on Machine Learning (ICML), 2014
    pdf bibtex

  • Bayesian Learning of Recursively Factored Environments
    Marc Bellemare, Joel Veness, Michael Bowling
    International Conference on Machine Learning (ICML), 2013
    pdf bibtex

  • Monte Carlo *-Minimax Search
    Marc Lanctot, Abdallah Saffidine, Joel Veness, Chris Archibald, Mark Winands
    International Joint Conference on Artificial Intelligence (IJCAI), 2013
    pdf arXive bibtex

  • Partition Tree Weighting
    Joel Veness, Martha White, Michael Bowling, András György
    Data Compression Conference (DCC), 2013
    pdf arXive bibtex

  • Sketch-based Linear Value Function Approximation
    Marc Bellemare, Joel Veness, Michael Bowling
    Neural Information Processing Systems (NIPS), 2012
    pdf bibtex

  • On Ensemble Techniques for AIXI Approximation
    Joel Veness, Peter Sunehag, Marcus Hutter
    Artificial General Intelligence (AGI), 2012
    pdf bibtex

  • Investigating Contingency Awareness Using Atari 2600 Games
    Marc Bellemare, Joel Veness, Michael Bowling
    Association for the Advancement of Artificial Intelligence (AAAI), 2012
    pdf bibtex

  • Context Tree Switching
    Joel Veness, Kee Siong Ng, Marcus Hutter, Michael Bowling
    Data Compression Conference (DCC), 2012
    pdf preprint source and binaries bibtex

  • Variance Reduction in Monte-Carlo Tree Search
    Joel Veness, Marc Lanctot, Michael Bowling
    Neural Information Processing Systems (NIPS), 2011
    pdf bibtex

  • An Approximation of the Universal Intelligence Measure
    Shane Legg, Joel Veness
    Ray Solomonoff 85th Memorial Conference, 2011
    pdf code arXive bibtex

  • Monte-Carlo Planning in Large POMDPs
    David Silver, Joel Veness
    Neural Information Processing Systems (NIPS), 2010
    pdf video and code bibtex

  • Reinforcement Learning via AIXI Approximation
    Joel Veness, Kee Siong Ng, Marcus Hutter, David Silver
    Association for the Advancement of Artificial Intelligence (AAAI), 2010
    pdf bibtex

  • Bootstrapping from Game Tree Search
    Joel Veness, David Silver, Will Uther, Alan Blair
    Neural Information Processing Systems (NIPS), 2009
    pdf video presentation bibtex

  • Effective Use of Transposition Tables in Stochastic Game Tree Search
    Joel Veness, Alan Blair
    IEEE Symposium on Computational Intelligence and Games (CIG), 2007
    pdf IEEE page bibtex

Technical Reports

  • Sparse Sequential Dirichlet Coding
    Joel Veness, Marcus Hutter
    Tech Report, 2012
    pdf arXive (obsoleted by the SAD estimator)

Theses

  • Approximate Universal Artificial Intelligence
    PhD Thesis, UNSW, 2011, pdf
  • Expectimax Enhancements for Stochastic Game Players
    Undergraduate Thesis, UNSW, 2006, pdf

 

Latest News

The journal version of the DQN paper is online.

Added our new AAAI'15 conference paper, a method for converting compression algorithms into RL agents.

Our 2011 JAIR paper "A Monte Carlo AIXI Approximation" was awarded an honorable mention for the 2014 IJCAI-JAIR best paper prize.