The Research Specialist will conduct independent and collaborative research in a computational cognitive neuroscience laboratory. The position focuses on mechanistic interpretability of large language models (LLMs) and other deep learning systems, with the goal of elucidating the internal mechanisms these models use to perform cognitive processes such as reasoning, abstraction, and symbolic computation. This role requires a demonstrated ability to design and execute research projects, analyze complex datasets, and contribute to peer-reviewed publications.
This is a one-year term position with the possibility of renewal contingent upon satisfactory performance and/or funding
• Collaborate with graduate students, postdoctoral researchers, and faculty within the lab and across affiliated research groups.
• Manage and organize research data, code repositories, and computational resources, including cloud-based GPU infrastructure.• Design and conduct mechanistic interpretability experiments on large language models, including probing, causal intervention, circuit-level analysis, and related techniques to identify and characterize the computational mechanisms underlying cognitive processes.
• Investigate how training conditions—including training data composition, data diversity, and model architecture—shape the emergence of symbolic and algorithmic mechanisms in neural networks.
• Extend mechanistic interpretability methods to new model classes, including but not limited to video models and reasoning-augmented language models.
• Develop and maintain research codebases for model training, interpretability analysis pipelines, and experimental workflows using Python and deep learning frameworks (e.g., PyTorch, JAX).
• Analyze experimental results using quantitative methods, including statistical analysis, visualization, and computational modeling.
• Collaborate with graduate students, postdoctoral researchers, and faculty within the lab and across affiliated research groups.
• Manage and organize research data, code repositories, and computational resources, including cloud-based GPU infrastructure.
• Prepare research findings for presentation at conferences and for publication in peer-reviewed journals; contribute as co-author on manuscripts.
• Conduct literature reviews and stay current with developments in mechanistic interpretability, cognitive science, and computational neuroscience.
• Assist with additional research tasks as directed by the Principal Investigator.
Essential Qualifications:
• Bachelor’s degree in computer science, cognitive science, neuroscience, mathematics, or a closely related field.
• Minimum of two years of research experience in machine learning, computational neuroscience, or a related area, including demonstrated contributions to research projects beyond coursework.
• Proficiency in Python and at least one deep learning framework (e.g., PyTorch, JAX, TensorFlow).
• Experience training and fine-tuning neural network models.
• Strong quantitative and analytical skills, including experience with statistical analysis and data visualization.
• Demonstrated ability to work independently on research tasks, including experimental design, data collection, and analysis.
• Excellent written and oral communication skills, with the ability to present technical findings to both specialist and general audiences.
• Co-authorship or significant contribution on at least one research paper or technical report.
• In-depth knowledge of modern deep learning architectures, particularly transformer-based language models.
• Understanding of experimental methodology in computational cognitive science or a related empirical discipline.
• Ability to manage multiple concurrent research projects and meet deadlines.
• Strong organizational skills for maintaining reproducible research workflows and well-documented codebases.
• Capacity to work effectively in a collaborative, interdisciplinary research environment
Preferred Qualifications:
• Experience with mechanistic interpretability methods (e.g., probing classifiers, activation patching, circuit analysis).
• Familiarity with the cognitive science or neuroscience of language, reasoning, or related cognitive processes.
• Experience with high-performance computing environments and cloud GPU platforms.
• Track record of peer-reviewed publication in machine learning, AI, or cognitive science venues.
This position is subject to the University's background check policy.
Princeton University is an Equal Opportunity Employer and all qualified applicants will receive consideration for employment without regard to age, race, color, religion, sex, sexual orientation, gender identity or expression, national origin, disability status, protected veteran status, or any other characteristic protected by law.
The University considers factors such as (but not limited to) scope and responsibilities of the position, candidate's qualifications, work experience, education/training, key skills, market, collective bargaining agreements as applicable, and organizational considerations when extending an offer. The posted salary range represents the University's good faith and reasonable estimate for a full-time position; salaries for part-time positions are pro-rated accordingly.
If the salary range on the posted position shows an hourly rate, this is the baseline; the actual hourly rate may be higher, depending on the position and factors listed above.
The University also offers a comprehensive benefit program to eligible employees. Please see this link for more information.