Education
| PhD in Informatics | June 2013 to May 2019 |
|---|---|
| Max Planck Institute for Mathematics in the Sciences | Leipzig, Germany |
- Supervisor: Professor Nihat Ay
- Dissertation Submitted: October 2017
- Dissertation Defended: May 2019
- Paternity leave: During my PhD I had my first and second child and took periods of leave for both.
| Master’s in Computational Neuroscience | October 2010 to May 2013 |
|---|---|
| Bernstein Centre for Computational Neuroscience Berlin | Berlin, Germany |
- Supervisor: Professor Manfred Opper
| Bachelor of Science | May 2004 to May 2009 |
|---|---|
| University of Toronto | Toronto, Canada |
- Specialist: Cognitive Science and Artificial Intelligence
- Major: Philosophy
- Minor: Statistics
Bibliography
Peer-Reviewed
- Sacha Sokoloski, Amir Aschner, and Ruben Coen-Cagli, “Modelling the Neural Code in Large Populations of Correlated Neurons,” ed. Jonathan W Pillow, Joshua I Gold, and Kenneth D Harris, eLife (2021). [Link] [PDF]
- Sacha Sokoloski, “Implementing a Bayes Filter in a Neural Circuit: The Case of Unknown Stimulus Dynamics,” Neural Computation (2017). [Link] [PDF]
Preprints
- Sacha Sokoloski and Philipp Berens, “Hierarchical Mixtures of Gaussians for Combined Dimensionality Reduction and Clustering” (arXiv, 2022). [Link] [PDF]
- Sacha Sokoloski, “A Biologically Realistic Model of Saccadic Eye Control with Probabilistic Population Codes,” arXiv:1210.4145 [Cs, Q-Bio] (2012). [Link] [PDF]
Theses
- Sacha Sokoloski, “Implementing Bayesian Inference with Neural Networks” (PhD thesis, University of Leipzig, 2019). [Link] [PDF]
- Sacha Sokoloski, “Efficient Stochastic Control with Kullback Leibler Costs Using Kernel Methods” (Master’s thesis, Technical University of Berlin, 2013). [PDF]
Professional Activities
Conferences and Workshops
| Population Models Workshop | July 2022 |
|---|---|
| University of Edinburgh | Edinburgh, Scotland |
- Invited Talk: Exact learning and inference in large-scale probabilistic graphical models
| Cosyne | March 2022 |
|---|---|
| Lisboa Congress Centre | Lisbon, Portugal |
- Poster: Disentangling neural dynamics with fluctuating hidden Markov models (Sacha Sokoloski, Ruben Coen-Cagli)
- Poster: An interpretable spline-LNP model to characterize feedforward and feedback processing in mouse dLGN (Lisa Schmors, Yannik Bauer, Ziwei Huang, Lukas Meyerolbersleben, Simon Renner, Ann H. Kotkat, Davide Crombie, Sacha Sokoloski, Laura Busse, Philipp Berens)
| Cosyne | February 2021 |
|---|---|
| Online |
- Poster: Information-limiting correlations linearize the neural code (Sacha Sokoloski, Ruben Coen-Cagli)
- Poster: Neural sampling from bimodal distributions in primary visual cortex (Ruben Coen-Cagli, Adam Kohn, Sacha Sokoloski)
| Cosyne | February 2020 |
|---|---|
| Hilton Denver City Center | Denver, USA |
- Poster: Mixture of poisson models for neural correlations and coding (Sacha Sokoloski, Ruben Coen-Cagli)
- Received travel grant
| Sense2Synapse | April 2017 |
|---|---|
| Rockefeller University | New York City, USA |
- Poster: Dynamic Bayesian Inference in the Brain
| International Conference on Mathematical Neuroscience | May 2016 |
|---|---|
| Antibes Juan-les-Pins Conference Centre | Juan-les-Pins, France |
- Talk: Learning to Filter Spike Trains when the Stimulus Dynamics are Unknown and Nonlinear
| HaL-10: Haskell in Leipzig | December 2015 |
|---|---|
| HTWK Leipzig | Leipzig, Germany |
- Talk: Typsichere Numerische Optimierung basierend auf Mannigfaltigkeiten
- English title: Typesafe Numerical Optimization based on Manifolds
Teaching and Mentorships
| Teaching Essential Statistics | Winter 2022/23 |
|---|---|
| University of Tübingen |
- Teaching the Master’s course “Essential Statistics for Neuroscience”.
| Google Summer of Code | Summer 2022 |
|---|---|
| Haskell.org |
- Mentored Zarak Mahmud through the Google Summer of Code program to add GPU computation to my Geometric Optimization Libraries.
| Master’s Student Supervision | Winter 2021/22 & Fall 2022 |
|---|---|
| University of Tübingen |
- Supervised Jure Majnik on developing code for the retinal-rl project.
| Google Summer of Code | Summer 2018 |
|---|---|
| Haskell.org |
- Mentored a student through the Google Summer of Code program on developing a library for parallel automatic differentiation for the Haskell.org organization.
- Project was overall successful.
Peer Review
- Neural Processing Letters (2022)
- Nature Neuroscience (Co-review, 2022)
- eLife (2021)
- Nature Communications (Co-review, 2019)
- IEEE Transactions on Neural Networks and Learning Systems (2017)
Employment History
Professional
| Postdoctoral Researcher | July 2021 to Present |
|---|---|
| University of Tübingen | Tübingen, Germany |
- Principal Investigator: Philipp Berens
- Project: Modelling neural circuits in the retina
| Research Fellow | October 2019 to April 2021 |
|---|---|
| Albert Einstein College of Medicine | New York, USA |
- Principal Investigator: Ruben Coen-Cagli
- Project: Modelling Bayesian inference in neural circuits and analyzing neural population response recordings
| Research Trainee | November 2017 to September 2019 |
|---|---|
| Albert Einstein College of Medicine | New York, USA |
- Principal Investigator: Ruben Coen-Cagli
Student
| Research Assistant | May 2012 to July 2012 |
|---|---|
| University of Geneva | Geneva, Switzerland |
- Principal Investigator: Professor Alexandre Pouget
- Project: Modelling sacaddic eye control using probabilistic population codes
| Programmer | August 2011 to May 2013 |
|---|---|
| Technical Unversity of Berlin | Berlin, Germany |
- Principal Investigator: Marianne Maertens
- Project: Developing software in Python to control scientific hardware for visual psychophysics
| Programmer | November 2009 to July 2010 |
|---|---|
| University of British Columbia | Vancouver, Canada |
- Principal Investigator: Professor Kurt Haas
- Project: Developing software in C to control a fast 2-photon microscope
| Research Assistant | April 2009 to July 2009 |
|---|---|
| University of Toronto | Toronto, Canada |
- Principal Investigator: Jun Luo
- Project: Writing an annotated bibliography in the Philosophy of Information & Computation
| Programmer | June 2008 to December 2008 |
|---|---|
| University of Toronto | Toronto, Canada |
- Principal Investigator: Professor Gerald Penn
- Project: Developing software in Prolog to estimate probability distributions over parsed linguistic corpora
| Research Assistant | June 2008 to December 2008 |
|---|---|
| University of Toronto | Toronto, Canada |
- Principal Investigator: Jun Luo
- Project: Coordinating the “Synesthetes” faculty meeting group, focusing on cognitive science and interdisciplinary research
| Programmer | December 2007 to August 2008 |
|---|---|
| University of Toronto | Toronto, Canada |
- Principal Investigator: Professor Fahiem Bacchus
- Project: Developing software in Java to schedule meeting times of potential graduate students
Miscellaneous
Summer/Winter Schools
| Dynamics of Multi-Level Systems | Summer 2015 |
|---|---|
| Max Planck Institute for the Physics of Complex Systems | Dresden, Germany |
| Autonomous Learning | Fall 2014 |
|---|---|
| Max Planck Institute for Mathematics in the Sciences | Leipzig, Germany |
| Deep Learning and Feature Learning | Summer 2012 |
|---|---|
| Institute for Pure and Applied Mathematics UCLA | Los Angeles, United States |
| The Future of the Embodied Mind | Fall 2011 |
|---|---|
| European Project: Extending Sensorimotor Contingencies to Cognition | San Sebastian, Spain |
| Advanced Numerical Bifurcation Analysis | Summer 2011 |
|---|---|
| Danish Centre for Applied Mathematics and Mechanics | Copenhagen, Denmark |
Scholarships and Awards
| Network of European Neuroscience Schools | Summer 2012 |
| Scholarship for training stay in lab of Alexandre Pouget | |
| Sunflower Scholarship in Philosophy | Spring 2007 |
| Award for undergraduate writing in philosophy |
Notable Skills
Languages
- English (Mother tongue)
- German (Fluent)
- French (Working knowledge)
Computer Skills
- Programming Languages: Haskell, C, Python, Matlab, R, Java, Prolog
- Advanced Linux user
- Experience maintaining a software library with approximately 10,000 lines of code
- Experience writing software for interfacing with scientific hardware