Intern - Machine Learning for Neuroscience

Research Intern – Machine Learning for Neuroscience

12 Month Duration - starting late fall 2026

The goal of the Neural Dynamics accelerator is to understand how the brain generates flexible behavior. We aim to uncover the algorithms—implemented by dynamics in brain-wide neural circuits—that allow animals to build internal models, process information, and choose actions.

We are searching for a graduate student intern who will analyze large-scale neural and behavioral recordings to uncover how populations of neurons coordinate to control movement.

This is a one-year, part-time research position at the intersection of machine learning and neuroscience. The intern will work with a rich existing dataset: two intermingled but genetically distinct populations of neurons in the striatum, recorded simultaneously, with individual cells tracked stably across many days, alongside a continuously measured motor output and synchronized muscle activity and video. The central question is a latent-variable modeling problem — how do the two populations jointly encode the motor output, and how much of their activity is shared between them versus private to each? No prior neuroscience background is required; the intern will be mentored directly. The position is a one-year commitment so that the intern has time to carry a project through to a scientific result, and strong work may contribute to a conference presentation or publication.

Essential Functions

• Characterize how coordinated activity across two simultaneously recorded neural populations encodes forelimb force
• Implement, fit, and evaluate machine learning and latent variable models that partition shared versus population-specific covariance, and benchmark them against recently published alternatives
• Relate neural population activity to behavioral variables using regression and classification methods, with appropriate cross-validation and controls
• Extend the analysis to electromyography (EMG) and synchronized behavioral video, and to how neural coding and behavior change across days of learning
• Produce documented, version-controlled Python analysis code that other team members can build on
• Communicate results to the mentoring scientists and the wider team, and contribute figures and methods text toward posters and manuscripts

Educational Objectives

• Hands-on experience applying machine learning and signal processing methods to large-scale neural and behavioral datasets
• Practical understanding of latent variable and dimensionality reduction models for multi-population data
• Experience with reproducible, collaborative scientific computing and open science practices
• Direct mentorship from staff scientists, exposure to systems neuroscience as a research area, and development of scientific writing and presentation skills

Required Education And Experience

• Currently enrolled in a master’s or PhD program in electrical and computer engineering, computer science, or a related quantitative field
• Completed coursework, prior to the start of the position, in programming, linear algebra, probability and statistics, and at least one of machine learning or signal processing
• Demonstrated experience with scientific computing in Python (numpy, scipy, pandas)
• Available to start in late October / early November 2026 and able to commit to the position for one year
• Able to work onsite at least one day per week

Preferred Education And Experience

• Coursework or project experience with linear regression and discrete classifiers (e.g., support vector machines)
• Coursework or project experience with dimensionality reduction and latent variable models (e.g., PCA, factor analysis, canonical correlation analysis)
• Experience with time series analysis and signal processing (filtering, spectral methods)
• Experience with scikit-learn and with deep learning frameworks such as PyTorch
• Experience with video-based pose estimation tools (e.g., DeepLabCut, SLEAP, Lightning Pose) or with biosignal analysis such as EMG
• Experience with software best practices (version control, code review, testing)
• Interest in computational or systems neuroscience; prior neuroscience coursework or research is welcome but not required
• Strong written and verbal communication skills, and the ability to work both independently and in a collaborative, multi-disciplinary environment