experience
Columbia University, Visual Inference Lab
Associate Research Scientist
February 2020 - now
- Leading a project to study and visualize how representations of images of objects change in neural networks and in human and non-human primate brain areas.
- Studying disentanglement of image sequences in variational autoencoders (VAE) and developing a feature stability-based inductive bias to make learning robust to adversarial attack.
- Developing a novel, brain-inspired neural network architecture based on biological feature map organization.
- Mentoring one PhD student. Mentored one visiting Master's student and served on their thesis defense committee.
Postdoctoral Research Scientist
February 2018 - February 2020
- Developed a novel, general, open-source visualization method for labeled data using hyperspheres, including statistical inference.
- Presented and published at 2 conferences.
New York University, Lab for Computational Vision & Visual Neuroscience Lab
Postdoctoral Research Associate
November 2016 - February 2018
- Developed a unifying theory of visual motion computation that reconciled decades of data with previous, separate theories.
- Fit hundreds of latent variable models (MATLAB) of visual motion computation to experiments involving 10,000-20,000 conditions per neuron with a high performance computing cluster.
- Presented and published at 1 conference.
Graduate Research Assistant
September 2010 - September 2016
- Designed novel stimuli using steerable pyramids and Fourier analysis to study visual motion selectivity in the brain in unprecedented detail.
- Led projects to measure neural responses to visual motion in behaving non-human primates.
- Presented and published at 4 conferences, and won a travel award.
University College London, Carandini Lab
Research Assistant
December 2008 - August 2010
- Implemented a breakthrough mouse behavioral training method for visual tasks and modeled their behavior in a highly cited publication.
- Developed a graphical interface and behavior logic program for technicians to analyze the behavioral performance and control the visual stimuli and behavioral reinforcement, in real time, of four mice simultaneously performing visual tasks.
- Developed interactive spike-sorting software for lab use.
- Studied thin- and thick-spiking neurons in Utah (10x10) electrode array recordings.
- Presented and published at 2 conferences.
Harvard Medical School, Born Lab
Research Assistant II
January 2007 - August 2008
- Built a relational database to standardize years of lab data.
- Managed the lab and its non-human primates, including behavioral training and surgery preparation and assistance.
Boston University, Computational Neuroscience
Master's Thesis Research
January 2006 - January 2007
- Developed a neural network model of visual motion computation in area MT resulting from end-stopping in primary visual cortex.
- Implemented various neural networks and machine learning-related models of vision, memory, and learning.
education
New York University
PhD, Computational Neuroscience (GPA 3.9/4.0)
September 2010 - September 2016
Boston University
MA, Cognitive & Neural Systems (GPA 3.9/4.0)
January 2006 - January 2007
Boston University
BA, Biology with specialization in Neuroscience, Minor in Mathematics Graduated Cum Laude
January 2006 - January 2007
- Activities:
- President, Treasurer of BU Outing Club.
- Vice President of Democracy Matters BU Chapter, a nonpartisan advocacy group for electoral campaign finance reform.
volunteer experience
Columbia University, Zuckerman Institute
December 2018 - now
- Organized and led institute town halls on postdoc-specific career resources and support.
New York Presbyterian Hospital and Columbia University
April 2020 - May 2020
- Distributed and inventoried scrubs amid staff shortages during the peak of the pandemic's first wave.
journal articles
Potnis AS, AD Zaharia, N Kriegeskorte (
AD Zaharia, AS Potnis, A Walther, N Kriegeskorte (
AD Zaharia, RLT Goris, JA Movshon, EP Simoncelli (2019). Compound stimuli reveal the structure of visual motion selectivity in macaque MT neurons.
B Vintch, AD Zaharia, JA Movshon, EP Simoncelli (2012). Efficient and direct estimation of a neural subunit model for sensory coding.
L Busse, A Ayaz, NT Dhruv, S Katzner, AB Saleem, ML Schölvinck, AD Zaharia, M Carandini (2011). The detection of visual contrast in the behaving mouse.
conferences
AD Zaharia, A Walther, N Kriegeskorte (2018). Visualizing the global geometry of population representations of multiple visual object categories with spheres.
AD Zaharia, RLT Goris, JA Movshon, EP Simoncelli (2017). Three-dimensional spatiotemporal receptive field structure in macaque area MT.
AD Zaharia, RLT Goris, JA Movshon, EP Simoncelli (2015). Compound stimuli reveal velocity separability of spatiotemporal receptive fields in macaque area MT.
AD Zaharia, RLT Goris, JA Movshon, EP Simoncelli (2014). Separability of spatiotemporal receptive field structure in macaque area MT.
AD Zaharia, B Vintch, JA Movshon, EP Simoncelli (2012). Characterizing MT Receptive Fields With Dense Filtered Noise.
B Vintch, AD Zaharia, JA Movshon, EP Simoncelli (2012). Fitting receptive fields in V1 and V2 as linear combinations of nonlinear subunits.
AD Zaharia, L Busse, S Katzner, M Carandini (2010). Visual psychometric functions in the mouse.
AD Zaharia, S Durand, L Busse, S Katzner, A Benucci, M Carandini (2009). Measuring the tuning of putative fast-spiking interneurons in cat area V1.
invited talks
AD Zaharia, RLT Goris, JA Movshon, EP Simoncelli (2017). Neural computation of visual motion in macaque area MT.
science communication
AD Zaharia (2020). Extracellular Recording.
skills
- Programming and data analysis: MATLAB, Python/NumPy/PyTorch, Mathematica, Java, Slurm/PBS, and Unix Shell scripting.
- Web technologies: HTML, CSS, PHP, Javascript, Processing, SQL, WordPress, and Amazon Web Services (AWS EC2).
- Applications: Adobe Illustrator and Photoshop, Microsoft, Apple, and Google word processing, spreadsheet, presentation, and database applications, and LaTeX.