Joel Z Leibo

jzleibo [at] mit.edu


Image of Joel Z. Leibo


 
I am a research scientist at Google DeepMind and a research affiliate with the McGovern Institute for Brain Research at MIT. My research is aimed at the following questions:
  • How do children learn to recognize objects by sight, and how can we build machines that do the same?
  • How does cortex support the computations underlying visual cognition?
  • How can we get deep reinforcement learning agents to perform complex cognitive behaviors?
  • How should we evaluate the performance of deep reinforcement learning agents?

Publications

  1. Perolat J*, Leibo JZ*, Zambaldi V, Beattie C, Tuyls K, and Graepel T.A multi-agent reinforcement learning model of common-pool resource appropriation. (2017). Advances in Neural Information Processing Systems (NIPS). Long Beach CA, USA (* = authors contributed equally) [bibtex]
  2. Hester T, Vecerik M, Pietquin O, Lanctot M, Schaul T, Piot B, Sendonaris A, Dulac-Arnold G, Osband I, Agapiou J, Leibo JZ, Gruslys A.Learning from Demonstrations for Real World Reinforcement Learning. (2017). arXiv preprint arXiv:1704.03732[bibtex]
  3. Sunehag P, Lever G, Gruslys A, Czarnecki WM, Zambaldi V, Jaderberg M, Lanctot M, Sonnerat N, Leibo JZ, Tuyls K, and Graepel T. Value-Decomposition Networks For Cooperative Multi-Agent Learning (2017). arXiv preprint arXiv:1706.05296[bibtex]
  4. Leibo JZ, Zambaldi V, Lanctot M, Marecki J, Graepel T. Multi-agent Reinforcement Learning in Sequential Social Dilemmas (2017). Proceedings of the 16th International Conference on Autonomous Agents and Multiagent Systems (AA-MAS 2017). Sao Paulo Brazil.  [bibtex] [blog post]
  5. Leibo JZ, Liao Q, Anselmi F, Freiwald WA, Poggio T.  View-Tolerant Face Recognition and Hebbian Learning Imply Mirror-Symmetric Neural Tuning to Head Orientation (2017). Current Biology 27, 1-6   , January 9, 2017. [bibtex] (press)
  6. Jaderberg M, Mnih V, Czarnecki WM, Schaul T, Leibo JZ, Silver D, Kavukcuoglu K. Reinforcement Learning with Unsupervised Auxiliary Tasks (2017).  International Conference on Learning Representations (ICLR). Toulon, France. [bibtex]
  7. Mutch J, Anselmi F, Tacchetti A, Rosasco L, Leibo JZ, Poggio T. Invariant Recognition Predicts Tuning of Neurons in Sensory Cortex (2017). Computational and Cognitive Neuroscience of Vision (Q. Zhao ed.) Springer Science+Business Media Singapore. DOI 10.1007/978-981-10-0213-7_5[bibtex]
  8. Beattie C, Leibo JZ, Teplyashin D, Ward T,  Wainwright M, Küttler H, Lefrancq A, Green S, Valdés V, Sadik A, Schrittwieser J, Anderson K, York S, Cant M, Cain A, Bolton A, Gaffney S, King H, Hassabis D, Legg S, Petersen S. DeepMind Lab (2016). arXiv:1612.03801 [cs.AI].  [bibtex] [blog post]
  9. Wang JX, Kurth-Nelson Z, Tirumala D, Soyer H, Leibo JZ, Munos R, Blundell C, Kumaran D, Botvinick M.  Learning to reinforcement learn (2016).  arXiv:1606.04460 [cs.LG]. [bibtex]
  10. Ba J, Hinton G, Mnih V, Leibo JZ, Ionescu C.  Using Fast Weights to Attend to the Recent Past (2016).  Advances in Neural Information Processing Systems (NIPS). Barcelona, Spain. [bibtex]
  11. Blundell C, Uria B, Pritzel A, Li Y, Ruderman A, Leibo, JZ, Rae JW, Wierstra D, Hassabis D.  Model-Free Episodic Control (2016).  arXiv:1606.04460 [stat.ML]. [bibtex]
  12. Liao Q, Leibo JZ, Poggio T.  How Important is Weight Symmetry in Backpropagation? (2016).   Proceedings of the thirtieth Association for the Advancement of Artificial Intelligence conference (AAAI). Phoenix, AZ. [bibtex]
  13. Leibo JZ, Cornebise J, Gomez S, Hassabis D.  Approximate Hubel-Wiesel Modules and the Data Structures of Neural Computation (2015).  arXiv:1512.08457 [cs.NE]. [bibtex]
  14. Leibo JZ, Liao Q, Anselmi F, Poggio T. The invariance hypothesis implies domain-specific regions in visual cortex (2015).  PLoS Computational Biology vol. 11. no. 10. e1004390 [bibtex]
  15. Anselmi F, Leibo JZ, Rosasco L, Mutch J, Tacchetti A, Poggio T. Unsupervised Learning of Invariant Representations (2015).  Theoretical Computer Science doi:10.1016/j.tcs.2015.06.048
  16. Liao Q, Leibo JZ and Tomaso Poggio T. Unsupervised learning of clutter-resistant visual representations from natural videos (2014). CBMM Memo No. 023. arXiv:1409.3879v1 . September 2014 [bibtex]
  17. Leibo JZ, Liao Q, Poggio T.  Subtasks of Unconstrained Face Recognition (2014).   9th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications. (VISAPP). Lisbon, Portugal. [bibtex]--- VISAPP presentation here --- Dataset available here
  18. Isik L, Meyers EM, Leibo JZ, Poggio T. The dynamics of invariant object recognition in the human visual system (2014). Journal of Neurophysiology vol. 111 no. 1 91-102 [bibtex]
  19. Liao Q, Leibo JZ, Mroueh Y, Poggio T. Can a biologically-plausible hierarchy effectively replace face detection, alignment, and recognition pipelines? (2013) arXiv:1311.4082, March 26, 2014.  [bibtex]
  20. Anselmi F, Leibo JZ, Rosasco L, Mutch J, Tacchetti A, Poggio T. Unsupervised Learning of Invariant Representations in Hierarchical Architectures (2013). arXiv:1311.4158, November 17, 2013. [bibtex]
  21. Liao Q, Leibo JZ, Poggio T.  Learning invariant representations and applications to face verification (2013).  Advances in Neural Information Processing Systems (NIPS). Lake Tahoe, NV. [bibtex]
  22. Leibo JZ. The Invariance Hypothesis and the Ventral Stream (2013).  MIT PhD thesis. [bibtex]
  23. Kim H,  Wohlwend J,  Leibo JZ,  Poggio T.  Body-form and body-pose recognition with a hierarchical model of the ventral stream (2013).    MIT-CSAIL-TR-2013-013,  CBCL-312. [bibtex]
  24. Poggio T, Mutch J, Anselmi F, Rosasco L, Leibo JZ, Tacchetti A.  The computational magic of the ventral stream: sketch of a theory (and why some deep architectures work) (2012).   MIT-CSAIL-TR-2012-035.
  25. Isik L*, Leibo JZ*, Poggio T.  Learning and disrupting invariance in visual recognition with a temporal association rule (2012).   Front. Comput. Neurosci. 6:37. doi: 10.3389/fncom.2012.00037 (* = authors contributed equally). Commentary by Bart and Hegdé: here[bibtex]
  26. Tan C, Leibo JZ, Poggio T.  Throwing Down the Visual Intelligence Gauntlet (2012).   Machine Learning for Computer Vision; eds: Cipolla R, Battiato S, Farinella GM. Springer: Studies in Computational Intelligence Vol. 411. [bibtex]
  27. Leibo JZ, Mutch J, Poggio T.  Why The Brain Separates Face Recognition From Object Recognition (2011).  Advances in Neural Information Processing Systems (NIPS). Granada, Spain.  [bibtex]
  28. Isik L, Leibo JZ, Mutch J, Lee SW, Poggio T.  A hierarchical model of peripheral vision (2011).  MIT-CSAIL-TR-2011-031, CBCL-300. [bibtex]
  29. Mutch J, Leibo JZ, Smale S, Rosasco L, Poggio T.  Neurons that confuse mirror-symmetric object views (2010).  MIT-CSAIL-TR-2010-062, CBCL-295.  [bibtex]
  30. Leibo JZ, Mutch J, Rosasco L, Ullman S, Poggio T.  Learning generic invariances in object recognition: translation and scale (2010). MIT-CSAIL-TR-2010-061, CBCL-294. [bibtex]
  31. Leibo JZ, Mutch J, Ullman S, Poggio T.  From primal templates to invariant recognition (2010).MIT-CSAIL-TR-2010-057, CBCL-293. [bibtex]

Selected conference abstracts

  1. Liao Q, Leibo JZ, Poggio T. Invariant Face Recognition in the Presence of Clutter (2014). Society for Neuroscience (823.20/II15). Washington DC.
  2. Leibo JZ, Anselmi F, Mutch J, Ebihara AF, Freiwald W, Poggio T. View-invariance and mirror-symmetric tuning in a model of the macaque face-processing system (2013). Computational and Systems Neuroscience (I-54). Salt Lake City, UT.
  3. Isik L, Meyers EM, Leibo JZ, Poggio T. Timing of invariant object recognition in the human visual system (2013). Computational and Systems Neuroscience (II-51). Salt Lake City, UT.
  4. Leibo JZ, Mutch J, Poggio T. Invariance to learned transformations in the ventral stream from V1 to AM (2012). Society for Neuroscience (263.05/Y14). New Orleans LA.
  5. Isik L, Meyers EM, Leibo JZ, Poggio T. Detecting invariant visual signals with MEG decoding (2012). Society for Neuroscience (262.14/X15). New Orleans LA.
  6. Kim H, Leibo JZ, Poggio T. Pose estimation and pose-invariant recognition with an extended hierarchical model of the ventral stream (2012). Society for Neuroscience (263.01/Y10). New Orleans LA.
  7. Ko EY, Leibo JZ, Poggio T. A hierarchical model of perspective-invariant scene identification (2011). Society for Neuroscience (486.16/OO26). Washington DC.
  8. Isik L, Leibo JZ, Lee S, Mutch J, Poggio T.  A hierarchical model of peripheral vision (2011). Society for Neuroscience (485.24/OO8). Washington DC.
  9. Leibo JZ, Mutch J, Poggio T.  Learning to discount transformations as the computational goal of visual cortex (2011). Presented at FGVC/CVPR 2011. Colorado Springs, CO. Available from Nature Precedings at dx.doi.org/10.1038/npre.2011.6078.1
  10. Lee SW, Leibo JZ, Mutch J, Poggio T.  Tuning errors and pooling errors in hierarchical models of object recognition (2011).  Computational and Systems Neuroscience (III-11).  Salt Lake City, UT.
  11. Leibo JZ, Mutch J, Poggio T.  How can cells in the anterior medial face patch be viewpoint invariant? (2011). Computational and Systems Neuroscience (III-10).  Salt Lake City UT. Availabe from Nature Precedings at dx.doi.org/10.1038/npre.2011.5845.1
  12. Leibo JZ, Mutch J, Rosasco L, Ullman S, Poggio T. The puzzle of initial invariance in object recognition (2010). Society for Neuroscience (675.2/LL10). San Diego CA.