2026 Applications are now open!

While applications will remain open until the positions are filled, applications received by June 26 will receive priority consideration. We anticipate fellowships will start in fall 2026, with some flexibility. All roles require on-site work in Seattle.

The UW Computational Neuroscience Center and the Shanahan Fellowship program have an exciting paid opportunity for pre-doctoral researchers at the interface of data and neuroscience. These roles are open to both current undergraduate and post-baccalaureate students.

Student fellows will be mentored by a current Shanahan Fellow, and they will gain hands-on research experience in neural computation, neural networks, and computational modeling/method development. Students will join a vibrant interdisciplinary research community with the opportunity to work with researchers at all levels at University of Washington and Allen Institute. As appropriate, fellows will be supported in opportunities to present and publish their work.

Please join our mailing list for updates from the program here. 

Dr. Janne Lappalainen

Bio:

Janne Lappalainen joined the Allen Institute and the University of Washington as a Shanahan Fellow in January 2026. He earned his B.Sc. in Physics from the University of Göttingen, his M.Sc. in Neuroengineering from the Technical University of Munich, and his Ph.D. in Computer Science from the University of Tübingen, supervised by Prof. Jakob Macke and Dr. Srinivas Turaga. His primary doctoral work was the development of deep mechanistic networks to study when and how detailed measurements of brain wiring can enable accurate, neuron-level predictions of neural dynamics. He also worked as a researcher at HHMI Janelia. As a Shanahan Fellow, Janne develops neural network simulations that integrate biological data with modern machine learning and theory, aiming toward accurate brain and organism-level models that support neuroscientific discovery.

Project: Theoretical models of learning in networks with synaptic plasticity typically assume homogeneous populations of neurons with unconstrained connectivity, neglecting cell types, sparse wiring, and regionally distinct activity patterns observed across the brain. In this project, at the intersection of theoretical neuroscience and machine learning, we constrain plastic networks with biologically grounded structure, such as distinct cell types, sparse connectivity, and oscillatory inputs as proxies for regional brain function. Preliminary experiments suggest that such constraints can improve trainability and task performance in comparison to various sequence-modeling baselines, particularly on tasks requiring longer temporal dependencies. We will investigate these effects systematically and characterize how learning mechanisms in structured networks differ from those in conventional homogeneous models, aiming to generate testable hypotheses about how biological structure shapes learning in the brain.

Mentorship and supervision will be tailored to the candidate’s experience and goals. An existing code repository can be extended with new models, machine learning tasks, and analysis of the resulting learning dynamics, offering collaboration also on software development. This position offers the opportunity to publish as first-author at a conference and to connect with the broader community of scientists at the Allen Institute and UW in preparation for graduate school or a research career.

Preferred skills: Mathematical foundations (linear algebra, dynamical systems, differential equations). Relevant background in sequence modeling ML, computational neuroscience, or physics. Proficiency in Python with experience in PyTorch or JAX. Experience with software version control like git / GitHub and agentic coding tools.

Dr. CiCi Zheng

Bio: CiCi Zheng is a Shanahan Foundation Fellow at the Allen Institute and the University of Washington, where she develops computational and theoretical tools to understand how cortical cell types are established during development, how that complexity supports learning, and how the structure of experience shapes learning in brains and machines. She completed her Ph.D. in Quantitative Biology at Cold Spring Harbor Laboratory, advised by Saket Navlakha and Alexei Koulakov, with a thesis on statistical modeling of networks in natural systems. Before that, she studied applied mathematics and biology at the College of William and Mary.

Project topics:

I) Data-driven method development. The mature brain is built from many specialized cell types. How are these types established during development, and how are they sustained for function in the adult? We work on two complementary lines of methods. The first uses neural optimal transport to infer developmental trajectories from multi-modal omics, testing whether modalities such as epigenetic state carry predictive information that sharpens our map of how neurons differentiate; early results from cortical Multiome data suggest they do. The second uses dynamics-aware unsupervised embeddings (anisotropic diffusion maps) to characterize cell states and their dynamics in the adult brain, recovering both stable states (e.g., SST interneurons) and the principal axes of motion in populations with persistent flux (e.g., cycling oligodendrocyte progenitors).

Data to interact with:

• Allen Mouse Whole Brain Atlas (ABC Atlas)
• Allen developing mouse visual cortex atlas (Example data access)
• Other high-quality publicly available datasets, e.g., from CZI CELLxGENE
• Potential other multi-modality omics or co-registered data from the Allen Institute

Example questions:

1. Do dynamics-principled embeddings recover functionally meaningful cell types?
2. Can we extract principal axes of dynamics, such as cycling and aging, in specific cell populations?
3. Does the epigenetic landscape carry predictive information about developmental trajectories?

Depending on interest, the project can lean more analysis-focused or modeling-focused.

II) Theory-oriented modeling and analysis. Where the first direction asks how the brain is built, this one asks how it learns. Starting from learning theory, we study how knowledge accrues when stimuli arrive sequentially: how signals like prediction error and novelty reshape what is learned from a new task, what is retained from old ones, and what transfers between them. From there, we ask whether those insights carry over to richer settings, either more biologically detailed cortical models or deep neural networks.

Example questions:

1. How gated plasticity in the cortex implement learning (computational neuroscience style)
2. Are there data selection strategies that enable efficient and robust learning? (deep learning)

Depending on interest, the work can lean more neuroscience-focused or ML-focused.

Preferences: Candidates with a bachelor’s degree in a quantitative field (mathematics, computer science, physics, or data science); some fluency in linear algebra, calculus, and dynamical systems; a curiosity to apply those skills to neuroscience or biology; and experience working with data in Python, including some familiarity with packages such as SciPy, scikit-learn, PyTorch, or JAX.

Dr. Libby Zhang

Bio: Libby Zhang develops computational and mathematical models to study how animals learn, adapt, and interact with their environment. She received her Ph.D. in electrical engineering from Stanford University, where she was advised by Dr. Scott Linderman and developed scalable statistical methods for quantifying behavior at various spatiotemporal scales. Prior to that, she received her M.Eng and B.S. in electrical engineering from MIT.

Description
The brain plays a central role in helping us adapt to and learn from a constantly changing world. Behavior is the brain’s primary interface with that world, and it provides a uniquely informative window into the brain’s underlying computational and neural mechanisms. Advances in quantifying complex behavior — at higher resolutions, across different timescales, and a much greater number of contexts — have opened new avenues into asking and understanding how the brain learns, adapts, and acts.

We are seeking a research assistant to contribute to a project in the topics areas of:

• Physics of behavior, such as movement composition and coupled interactive dynamics
• Learning and decision making across timescales
• Social communication and information flow