High-Dimensional Dynamics & Computation Group
David G. Clark
Group Leader & Associate Research Scientist
Flatiron Institute, Center for Computational Neuroscience
Simons Foundation
While I am highly enthusiastic about AI in the theoretical sciences, please refrain from using it in emails to me; I will do the same. Thank you!
I seek a conceptual understanding of how large neural circuits process information. To this end, I develop theories linking synaptic connectivity, neuronal dynamics, and computation. Concretely, I use tools from statistical physics and machine learning to analyze high-dimensional nonlinear network models and connect them to experimental data.
Key directions include how task demands sculpt network dynamics; the origins and dynamics of structured representations, such as manifolds and associative memories; novel forms of processing enabled by synaptic plasticity; and theory-driven interpretation of large-scale population recordings.
Multiple postdoc positions are available in my group through the Flatiron Research Fellowship. Feel free to email me with any questions.
Ph.D. students can work with me through Columbia University.
From 2025 to 2026, I was a Research Fellow at the Kempner Institute for the Study of Natural and Artificial Intelligence at Harvard University. I received a Ph.D. in Neurobiology and Behavior from Columbia University in 2025, where I was advised by Larry Abbott and worked with Ashok Litwin-Kumar, Haim Sompolinsky, and Richard Axel. Before that, I studied physics and computer science at UC Berkeley, where I did research with Kristofer Bouchard.
My publications are listed below, or see my Google Scholar. My CV is here.
Outside of science, I see a lot of Broadway.
Publications & preprints
- (2026). Associative synaptic plasticity creates dynamic persistent activity. PNAS. JournalPDFbioRxiv
- (2026). Transient dynamics of associative memory models. Physical Review E. JournalPDFarXiv
- (2026). Linear equivalence of nonlinear recurrent neural networks. arXiv. arXiv
- (2026). Structure, disorder, and dynamics in task-trained recurrent neural circuits. bioRxiv. bioRxivKempner blogCode
- (2025). A theory of multi-task computation and task selection. bioRxiv. bioRxiv
- (2025). Connectivity structure and dynamics of nonlinear recurrent neural networks. Physical Review X. JournalPDFarXiv
- (2025). Simplified derivations for high-dimensional convex learning problems. SciPost Physics Lecture Notes. JournalPDFarXiv
- (2025). Theories of structure, dynamics, and plasticity in neural circuits. Ph.D. thesis, Columbia University. Academic Commons
- (2025). Structure of activity in multiregion recurrent neural networks. PNAS. JournalPDFarXiv
- (2025). Symmetries and continuous attractors in disordered neural circuits. bioRxiv. bioRxiv
- (2024). Theory of coupled neuronal-synaptic dynamics. Physical Review X. JournalPDFarXivViewpointCode
- (2023). Dimension of activity in random neural networks. Physical Review Letters. JournalPDFarXiv
- (2021). Olfactory landmarks and path integration converge to form a cognitive spatial map. Neuron. JournalPDFVideoCode
- (2021). Credit assignment through broadcasting a global error vector. NeurIPS 2021. arXivCode
- (2019). Unsupervised discovery of temporal structure in noisy data with dynamical components analysis. NeurIPS 2019. arXivCode
- (2017). Neuromorphic Kalman filter implementation in IBM's TrueNorth. Journal of Physics: Conference Series. JournalPDF
Invited talks
- Oct 2026SIAM NY-NJ-PA Section Conference minisymposium on stochastic neuroscience, Rutgers University, New Brunswick
- Oct 2026Dynamics in Neuroscience Workshop, National Institute for Theory and Mathematics in Biology, Chicago
- Sep 2026Bernstein Conference workshop: “Quo vadis, neural network theory?” Frankfurt
- Jun 2026ICMNS mini-symposium: “Recent Advances in the Study of Disorder in Recurrent Neural Networks,” Montreal
- Mar 2026Dynamical Systems Seminar, Department of Mathematics and Statistics, Boston University
- Feb 2026Computational Neuroscience Seminar, Courant Institute, New York University
- Jan 2026van Vreeswijk Theoretical Neuroscience Seminar (virtual)
- Jan 2026Theoretical Physics for Artificial Intelligence, Aspen Center for Physics
- Nov 2025Simons Collaboration on the Physics of Learning and Neural Computation, kickoff workshop, Stanford University
- Jun 2025Gatsby Tri-Centre Meeting, University College London
- Jun 2025Shervin Safavi group, TU Dresden (virtual)
- Mar 2025CoSyNe workshop: “Collectively Emerged Timescales,” Montreal
- Sep 2024Cengiz Pehlevan group, Harvard University
- Sep 2024Xiao-Jing Wang group, New York University
- May 2024Youth in High Dimensions, International Centre for Theoretical Physics, Trieste
- May 2024Hakan Türeci group, Princeton University
- Feb 2024University of Washington Theoretical Neuroscience Journal Club (virtual)
- Dec 2023Rutgers 125th Statistical Mechanics Conference
- Sep 2023Bernstein Conference workshop: “Relationship Between Multi-level Network Connectivity and Neural Dynamics,” Berlin
- Jun 2023Junior Theoretical Neuroscientist Workshop, Flatiron Center for Computational Neuroscience
- Apr 2023Ilya Nemenman lab, Emory University
- Apr 2023Theoretical Neuroscience Journal Club organized by Xaq Pitkow (virtual)
- Feb 2023Wulfram Gerstner group, EPFL
- Feb 2023Les Houches workshop: “Toward a Theory of Artificial and Biological Neural Networks”
- Oct 2022Center for the Physics of Biological Function, Princeton University
- Sep 2022Redwood Center for Theoretical Neuroscience, UC Berkeley