Jonathan Binas
Jonathan has a background in Physics, has worked in Neuroscience, Electronics (developed analog hardware to run neural networks), and is currently a postdoc with Yoshua Bengio, working on deep and reinforcement learning problems. His current interests include modular policies, brain-inspired learning algorithms, and alternative computing substrates.
Recent publications
-
Reinforcement Learning with Random Delays.
Ramstedt, Simon; Bouteiller, Yann; Beltrame, Giovanni; Pal, Christopher; Binas, Jonathan.
arXiv preprint arXiv:2010.02966.
2020.
https://arxiv.org/abs/2010.02966
-
DDD20 End-to-End Event Camera Driving Dataset: Fusing Frames and Events with Deep Learning for Improved Steering Prediction.
Hu, Yuhuang; Binas, Jonathan; Neil, Daniel; Liu, Shih-Chii; Delbruck, Tobi.
ITSC.
2020.
https://arxiv.org/abs/2005.08605
-
Out-of-distribution generalization via risk extrapolation (rex).
Krueger, David; Caballero, Ethan; Jacobsen, Joern-Henrik; Zhang, Amy; Binas, Jonathan; Priol, Remi Le; Courville, Aaron.
arXiv preprint arXiv:2003.00688.
2020.
https://arxiv.org/abs/2003.00688
-
Reinforcement learning with competitive ensembles of information-constrained primitives.
Goyal, Anirudh; Sodhani, Shagun; Binas, Jonathan; Peng, Xue Bin; Levine, Sergey; Bengio, Yoshua.
ICLR.
2020.
https://openreview.net/forum?id=ryxgJTEYDr
-
The Journey is the Reward: Unsupervised Learning of Influential Trajectories.
Binas, Jonathan; Ozair, Sherjil; Bengio, Yoshua.
ICML Workshop: ERL.
2019.
https://arxiv.org/abs/1905.09334
-
Retrieving Signals with Deep Complex Extractors.
Trabelsi, Chiheb; Bilaniuk, Olexa; Dia, Ousmane; Zhang, Ying; Ravanelli, Mirco; Binas, Jonathan; Rostamzadeh, Negar; Pal, Christopher J.
NeurIPS Workshop: Deep Inverse Models.
2019.
https://openreview.net/forum?id=H1x22Xn5Ur
-
Reinforcement Learning for Sustainable Agriculture.
Binas, Jonathan; Luginbuehl, Leonie; Bengio, Yoshua.
ICML Climate Change Workshop.
2019.
https://www.climatechange.ai/CameraReady/40/CameraReadySubmission/Yield_optimization.pdf
-
Lagrangian dynamics of dendritic microcircuits enables real-time backpropagation of errors.
Dold, Dominik; Kungl, Akos F; Sacramento, João; Petrovici, Mihai A; Schindler, Kaspar; Binas, Jonathan; Bengio, Yoshua; Senn, Walter.
Cosyne.
2019.
http://www.kip.uni-heidelberg.de/Veroeffentlichungen/download.php/6269/temp/3855.pdf
-
State-reification networks: Improving generalization by modeling the distribution of hidden representations.
Lamb, Alex; Binas, Jonathan; Goyal, Anirudh; Subramanian, Sandeep; Mitliagkas, Ioannis; Kazakov, Denis; Bengio, Yoshua; Mozer, Michael C.
ICML.
2019.
http://proceedings.mlr.press/v97/lamb19a.html
-
Analogue electronic neural network.
Binas, Jonathan; Neil, Daniel.
US Patent App. 16/078,769.
2019.
https://patents.google.com/patent/US20190050720A1/en
-
Fortified networks: Improving the robustness of deep networks by modeling the manifold of hidden representations.
Lamb, Alex; Binas, Jonathan; Goyal, Anirudh; Serdyuk, Dmitriy; Subramanian, Sandeep; Mitliagkas, Ioannis; Bengio, Yoshua.
arXiv preprint arXiv:1804.02485.
2018.
https://arxiv.org/abs/1804.02485
-
Generalization of equilibrium propagation to vector field dynamics.
Scellier, Benjamin; Goyal, Anirudh; Binas, Jonathan; Mesnard, Thomas; Bengio, Yoshua.
ICLR Workshop.
2018.
https://arxiv.org/abs/1808.04873
-
Fully discretized training of neural networks through direct feedback.
Mesnard, Thomas; Vignoud, Gaëtan; Binas, Jonathan; Bengio, Yoshua.
preprint.
2018.
-
Low-memory convolutional neural networks through incremental depth-first processing.
Binas, Jonathan; Bengio, Yoshua.
arXiv preprint arXiv:1804.10727.
2018.
https://arxiv.org/abs/1804.10727
-
Analog electronic deep networks for fast and efficient inference.
Binas, Jonathan; Neil, Daniel; Indiveri, Giacomo; Liu, Shih-Chii; Pfeiffer, Michael.
Proc Conf. Syst. Mach. Learning (SysML).
2018.
https://mlsys.org/Conferences/2019/doc/2018/179.pdf
-
Sparse attentive backtracking: Temporal credit assignment through reminding.
Ke, Nan Rosemary; Goyal, Anirudh; Bilaniuk, Olexa; Binas, Jonathan; Mozer, Michael C; Pal, Chris; Bengio, Yoshua.
NeurIPS.
2018.
http://papers.nips.cc/paper/7991-sparse-attentive-backtracking-temporal-credit-assignment-through-reminding
-
Brain-inspired models and systems for distributed computation.
Binas, Jonathan.
Thesis; ETH Zurich.
2017.
https://www.research-collection.ethz.ch/bitstream/handle/20.500.11850/216012/1/thesis.pdf
-
DDD17: End-to-end DAVIS driving dataset.
Binas, Jonathan; Neil, Daniel; Liu, Shih-Chii; Delbruck, Tobi.
ICML Workshop: ITS.
2017.
https://arxiv.org/abs/1711.01458
-
Precise neural network computation with imprecise analog devices.
Binas, Jonathan; Neil, Daniel; Indiveri, Giacomo; Liu, Shih-Chii; Pfeiffer, Michael.
arXiv preprint arXiv:1606.07786.
2016.
https://arxiv.org/abs/1606.07786
-
Deep counter networks for asynchronous event-based processing.
Binas, Jonathan; Indiveri, Giacomo; Pfeiffer, Michael.
arXiv preprint arXiv:1611.00710.
2016.
https://arxiv.org/abs/1611.00710
-
Spiking analog vlsi neuron assemblies as constraint satisfaction problem solvers.
Binas, Jonathan; Indiveri, Giacomo; Pfeiffer, Michael.
ISCAS.
2016.
https://ieeexplore.ieee.org/abstract/document/7538992/
-
Local structure supports learning of deterministic behavior in recurrent neural networks.
Binas, Jonathan; Indiveri, Giacomo; Pfeiffer, Michael.
BMC Neuroscience.
2015.
https://link.springer.com/article/10.1186/1471-2202-16-S1-P195
-
Fast-classifying, high-accuracy spiking deep networks through weight and threshold balancing.
Diehl, Peter U; Neil, Daniel; Binas, Jonathan; Cook, Matthew; Liu, Shih-Chii; Pfeiffer, Michael.
IJCNN.
2015.
https://ieeexplore.ieee.org/abstract/document/7280696/
-
Local structure helps learning optimized automata in recurrent neural networks.
Binas, Jonathan; Indiveri, Giacomo; Pfeiffer, Michael.
IJCNN.
2015.
https://ieeexplore.ieee.org/abstract/document/7280714/
-
Learning and stabilization of winner-take-all dynamics through interacting excitatory and inhibitory plasticity.
Binas, Jonathan; Rutishauser, Ueli; Indiveri, Giacomo; Pfeiffer, Michael.
Frontiers in computational neuroscience.
2014.
https://www.frontiersin.org/articles/10.3389/fncom.2014.00068/full
-
Synthesizing cognition in neuromorphic electronic systems.
Neftci, Emre; Binas, Jonathan; Rutishauser, Ueli; Chicca, Elisabetta; Indiveri, Giacomo; Douglas, Rodney J.
Proceedings of the National Academy of Sciences (PNAS).
2013.
https://www.pnas.org/content/110/37/E3468.short
-
Systematic Construction of Finite State Automata Using VLSI Spiking Neurons.
Neftci, Emre; Binas, Jonathan; Chicca, Elisabetta; Indiveri, Giacomo; Douglas, Rodney.
Biomimetic and Biohybrid Systems.
2012.
https://link.springer.com/chapter/10.1007/978-3-642-31525-1_52
-
Linear and cyclic porphyrin hexamers as near-infrared emitters in organic light-emitting diodes.
Fenwick, Oliver; Sprafke, Johannes K; Binas, Jonathan; Kondratuk, Dmitry V; Di Stasio, Francesco; Anderson, Harry L; Cacialli, Franco.
Nano letters.
2011.
https://pubs.acs.org/doi/abs/10.1021/nl2008778
© J. Binas, 2020