Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox
Picture for David Balduzzi

David Balduzzi

Real World Games Look Like Spinning Tops


Apr 20, 2020
Wojciech Marian Czarnecki, Gauthier Gidel, Brendan Tracey, Karl Tuyls, Shayegan Omidshafiei, David Balduzzi, Max Jaderberg

Figure 1 for Real World Games Look Like Spinning Tops
Figure 2 for Real World Games Look Like Spinning Tops
Figure 3 for Real World Games Look Like Spinning Tops
Figure 4 for Real World Games Look Like Spinning Tops

  Access Paper or Ask Questions

Learning to Resolve Alliance Dilemmas in Many-Player Zero-Sum Games


Feb 27, 2020
Edward Hughes, Thomas W. Anthony, Tom Eccles, Joel Z. Leibo, David Balduzzi, Yoram Bachrach

Figure 1 for Learning to Resolve Alliance Dilemmas in Many-Player Zero-Sum Games
Figure 2 for Learning to Resolve Alliance Dilemmas in Many-Player Zero-Sum Games
Figure 3 for Learning to Resolve Alliance Dilemmas in Many-Player Zero-Sum Games
Figure 4 for Learning to Resolve Alliance Dilemmas in Many-Player Zero-Sum Games
* Accepted for publication at AAMAS 2020 
  Access Paper or Ask Questions

From Poincaré Recurrence to Convergence in Imperfect Information Games: Finding Equilibrium via Regularization


Feb 19, 2020
Julien Perolat, Remi Munos, Jean-Baptiste Lespiau, Shayegan Omidshafiei, Mark Rowland, Pedro Ortega, Neil Burch, Thomas Anthony, David Balduzzi, Bart De Vylder, Georgios Piliouras, Marc Lanctot, Karl Tuyls

Figure 1 for From Poincaré Recurrence to Convergence in Imperfect Information Games: Finding Equilibrium via Regularization
Figure 2 for From Poincaré Recurrence to Convergence in Imperfect Information Games: Finding Equilibrium via Regularization
Figure 3 for From Poincaré Recurrence to Convergence in Imperfect Information Games: Finding Equilibrium via Regularization
Figure 4 for From Poincaré Recurrence to Convergence in Imperfect Information Games: Finding Equilibrium via Regularization
* 43 pages 
  Access Paper or Ask Questions

Minimax Theorem for Latent Games or: How I Learned to Stop Worrying about Mixed-Nash and Love Neural Nets


Feb 14, 2020
Gauthier Gidel, David Balduzzi, Wojciech Marian Czarnecki, Marta Garnelo, Yoram Bachrach

Figure 1 for Minimax Theorem for Latent Games or: How I Learned to Stop Worrying about Mixed-Nash and Love Neural Nets
Figure 2 for Minimax Theorem for Latent Games or: How I Learned to Stop Worrying about Mixed-Nash and Love Neural Nets
Figure 3 for Minimax Theorem for Latent Games or: How I Learned to Stop Worrying about Mixed-Nash and Love Neural Nets
* 15 pages 
  Access Paper or Ask Questions

Smooth markets: A basic mechanism for organizing gradient-based learners


Jan 18, 2020
David Balduzzi, Wojciech M Czarnecki, Thomas W Anthony, Ian M Gemp, Edward Hughes, Joel Z Leibo, Georgios Piliouras, Thore Graepel

Figure 1 for Smooth markets: A basic mechanism for organizing gradient-based learners
Figure 2 for Smooth markets: A basic mechanism for organizing gradient-based learners
Figure 3 for Smooth markets: A basic mechanism for organizing gradient-based learners
* ICLR 2020 
 * 18 pages, 3 figures 
  Access Paper or Ask Questions

LOGAN: Latent Optimisation for Generative Adversarial Networks


Dec 02, 2019
Yan Wu, Jeff Donahue, David Balduzzi, Karen Simonyan, Timothy Lillicrap

Figure 1 for LOGAN: Latent Optimisation for Generative Adversarial Networks
Figure 2 for LOGAN: Latent Optimisation for Generative Adversarial Networks
Figure 3 for LOGAN: Latent Optimisation for Generative Adversarial Networks
Figure 4 for LOGAN: Latent Optimisation for Generative Adversarial Networks

  Access Paper or Ask Questions

Differentiable Game Mechanics


May 13, 2019
Alistair Letcher, David Balduzzi, Sebastien Racaniere, James Martens, Jakob Foerster, Karl Tuyls, Thore Graepel

Figure 1 for Differentiable Game Mechanics
Figure 2 for Differentiable Game Mechanics
Figure 3 for Differentiable Game Mechanics
Figure 4 for Differentiable Game Mechanics
* Journal of Machine Learning Research (JMLR), v20 (84) 1-40, 2019 
 * JMLR 2019, journal version of arXiv:1802.05642 
  Access Paper or Ask Questions

Open-ended Learning in Symmetric Zero-sum Games


Jan 23, 2019
David Balduzzi, Marta Garnelo, Yoram Bachrach, Wojciech M. Czarnecki, Julien Perolat, Max Jaderberg, Thore Graepel

Figure 1 for Open-ended Learning in Symmetric Zero-sum Games
Figure 2 for Open-ended Learning in Symmetric Zero-sum Games
Figure 3 for Open-ended Learning in Symmetric Zero-sum Games
Figure 4 for Open-ended Learning in Symmetric Zero-sum Games
* 18 pages, 7 figures 
  Access Paper or Ask Questions

Stable Opponent Shaping in Differentiable Games


Nov 20, 2018
Alistair Letcher, Jakob Foerster, David Balduzzi, Tim Rocktäschel, Shimon Whiteson

Figure 1 for Stable Opponent Shaping in Differentiable Games
Figure 2 for Stable Opponent Shaping in Differentiable Games
Figure 3 for Stable Opponent Shaping in Differentiable Games
Figure 4 for Stable Opponent Shaping in Differentiable Games
* 20 pages, 7 figures 
  Access Paper or Ask Questions

Re-evaluating Evaluation


Oct 30, 2018
David Balduzzi, Karl Tuyls, Julien Perolat, Thore Graepel

Figure 1 for Re-evaluating Evaluation
Figure 2 for Re-evaluating Evaluation
Figure 3 for Re-evaluating Evaluation
Figure 4 for Re-evaluating Evaluation
* NIPS 2018, final version 
  Access Paper or Ask Questions

The Mechanics of n-Player Differentiable Games


Jun 06, 2018
David Balduzzi, Sebastien Racaniere, James Martens, Jakob Foerster, Karl Tuyls, Thore Graepel

Figure 1 for The Mechanics of n-Player Differentiable Games
Figure 2 for The Mechanics of n-Player Differentiable Games
Figure 3 for The Mechanics of n-Player Differentiable Games
Figure 4 for The Mechanics of n-Player Differentiable Games
* PMLR volume 80, 2018 
 * ICML 2018, final version 
  Access Paper or Ask Questions

Neural Taylor Approximations: Convergence and Exploration in Rectifier Networks


Jun 06, 2018
David Balduzzi, Brian McWilliams, Tony Butler-Yeoman

Figure 1 for Neural Taylor Approximations: Convergence and Exploration in Rectifier Networks
Figure 2 for Neural Taylor Approximations: Convergence and Exploration in Rectifier Networks
Figure 3 for Neural Taylor Approximations: Convergence and Exploration in Rectifier Networks
Figure 4 for Neural Taylor Approximations: Convergence and Exploration in Rectifier Networks
* PMLR volume 70, 2017 
 * ICML 2017, final version 
  Access Paper or Ask Questions

Strongly-Typed Agents are Guaranteed to Interact Safely


Jun 06, 2018
David Balduzzi

* PMLR volume 70, 2017 
 * ICML 2017, final version 
  Access Paper or Ask Questions

The Shattered Gradients Problem: If resnets are the answer, then what is the question?


Jun 06, 2018
David Balduzzi, Marcus Frean, Lennox Leary, JP Lewis, Kurt Wan-Duo Ma, Brian McWilliams

Figure 1 for The Shattered Gradients Problem: If resnets are the answer, then what is the question?
Figure 2 for The Shattered Gradients Problem: If resnets are the answer, then what is the question?
Figure 3 for The Shattered Gradients Problem: If resnets are the answer, then what is the question?
Figure 4 for The Shattered Gradients Problem: If resnets are the answer, then what is the question?
* PMLR volume 70 (2017) 
 * ICML 2017, final version 
  Access Paper or Ask Questions

Deep Reconstruction-Classification Networks for Unsupervised Domain Adaptation


Aug 01, 2016
Muhammad Ghifary, W. Bastiaan Kleijn, Mengjie Zhang, David Balduzzi, Wen Li

Figure 1 for Deep Reconstruction-Classification Networks for Unsupervised Domain Adaptation
Figure 2 for Deep Reconstruction-Classification Networks for Unsupervised Domain Adaptation
Figure 3 for Deep Reconstruction-Classification Networks for Unsupervised Domain Adaptation
Figure 4 for Deep Reconstruction-Classification Networks for Unsupervised Domain Adaptation
* to appear in European Conference on Computer Vision (ECCV) 2016 
  Access Paper or Ask Questions

Scatter Component Analysis: A Unified Framework for Domain Adaptation and Domain Generalization


Jul 26, 2016
Muhammad Ghifary, David Balduzzi, W. Bastiaan Kleijn, Mengjie Zhang

Figure 1 for Scatter Component Analysis: A Unified Framework for Domain Adaptation and Domain Generalization
Figure 2 for Scatter Component Analysis: A Unified Framework for Domain Adaptation and Domain Generalization
Figure 3 for Scatter Component Analysis: A Unified Framework for Domain Adaptation and Domain Generalization
Figure 4 for Scatter Component Analysis: A Unified Framework for Domain Adaptation and Domain Generalization
* to appear in IEEE Transactions on Pattern Analysis and Machine Intelligence 
  Access Paper or Ask Questions

Strongly-Typed Recurrent Neural Networks


May 24, 2016
David Balduzzi, Muhammad Ghifary

Figure 1 for Strongly-Typed Recurrent Neural Networks
Figure 2 for Strongly-Typed Recurrent Neural Networks
* 10 pages, final version, ICML 2016 
  Access Paper or Ask Questions

Deep Online Convex Optimization by Putting Forecaster to Sleep


Apr 08, 2016
David Balduzzi

* Rendered obsolete by arXiv:1604.01952. The new version contains the same basic results, with major changes to exposition and minor changes to terminology 
  Access Paper or Ask Questions

Deep Online Convex Optimization with Gated Games


Apr 07, 2016
David Balduzzi

* 13 pages. This paper renders arXiv:1509.01851 obsolete. It contains the same basic results, with major changes to exposition and minor changes to terminology 
  Access Paper or Ask Questions

Compliance-Aware Bandits


Feb 09, 2016
Nicolás Della Penna, Mark D. Reid, David Balduzzi

Figure 1 for Compliance-Aware Bandits
Figure 2 for Compliance-Aware Bandits
Figure 3 for Compliance-Aware Bandits

  Access Paper or Ask Questions

Semantics, Representations and Grammars for Deep Learning


Sep 29, 2015
David Balduzzi

* 20 pages, many diagrams 
  Access Paper or Ask Questions

Compatible Value Gradients for Reinforcement Learning of Continuous Deep Policies


Sep 10, 2015
David Balduzzi, Muhammad Ghifary

Figure 1 for Compatible Value Gradients for Reinforcement Learning of Continuous Deep Policies
Figure 2 for Compatible Value Gradients for Reinforcement Learning of Continuous Deep Policies
Figure 3 for Compatible Value Gradients for Reinforcement Learning of Continuous Deep Policies
* 27 pages 
  Access Paper or Ask Questions

Domain Generalization for Object Recognition with Multi-task Autoencoders


Aug 31, 2015
Muhammad Ghifary, W. Bastiaan Kleijn, Mengjie Zhang, David Balduzzi

Figure 1 for Domain Generalization for Object Recognition with Multi-task Autoencoders
Figure 2 for Domain Generalization for Object Recognition with Multi-task Autoencoders
Figure 3 for Domain Generalization for Object Recognition with Multi-task Autoencoders
Figure 4 for Domain Generalization for Object Recognition with Multi-task Autoencoders
* accepted in ICCV 2015 
  Access Paper or Ask Questions

Kickback cuts Backprop's red-tape: Biologically plausible credit assignment in neural networks


Nov 23, 2014
David Balduzzi, Hastagiri Vanchinathan, Joachim Buhmann

Figure 1 for Kickback cuts Backprop's red-tape: Biologically plausible credit assignment in neural networks
Figure 2 for Kickback cuts Backprop's red-tape: Biologically plausible credit assignment in neural networks
* 7 pages. To appear, AAAI-15 
  Access Paper or Ask Questions

Falsifiable implies Learnable


Aug 28, 2014
David Balduzzi


  Access Paper or Ask Questions

Cortical prediction markets


Jan 07, 2014
David Balduzzi

Figure 1 for Cortical prediction markets
Figure 2 for Cortical prediction markets
Figure 3 for Cortical prediction markets
* To appear, AAMAS 2014 
  Access Paper or Ask Questions

Correlated random features for fast semi-supervised learning


Nov 05, 2013
Brian McWilliams, David Balduzzi, Joachim M. Buhmann

Figure 1 for Correlated random features for fast semi-supervised learning
Figure 2 for Correlated random features for fast semi-supervised learning
Figure 3 for Correlated random features for fast semi-supervised learning
Figure 4 for Correlated random features for fast semi-supervised learning
* 15 pages, 3 figures, 6 tables 
  Access Paper or Ask Questions

Randomized co-training: from cortical neurons to machine learning and back again


Oct 24, 2013
David Balduzzi

Figure 1 for Randomized co-training: from cortical neurons to machine learning and back again
* NIPS workshop: Randomized methods for machine learning 
  Access Paper or Ask Questions

Metabolic cost as an organizing principle for cooperative learning


Feb 09, 2013
David Balduzzi, Pedro A Ortega, Michel Besserve

Figure 1 for Metabolic cost as an organizing principle for cooperative learning
Figure 2 for Metabolic cost as an organizing principle for cooperative learning
* 14 pages, 2 figures, to appear in Advances in Complex Systems 
  Access Paper or Ask Questions