Research Fellow

Mass General Brigham • Boston, MA

10h • Onsite

About The Position

The Research Fellow will contribute to an ongoing project to create, test, and validate AI methods that help clinicians detect errors (false outputs) generated by multimodal AI models. The current research focus uses mechanistic interpretability techniques in the natural language domain, including large language models (LLMs).

The fellow will work collaboratively with computer scientists and clinicians to help design, coordinate, carry out, and analyze domain-expert clinician annotations and assessments of mechanistic interpretability methods in clinical medicine. The Multi-Modal AI in Neuroimaging (MMAIN) Lab is a multi-institutional, multidisciplinary collaboration involving faculty from Mass General Brigham, Northeastern University's Khoury College of Computer Science, and Dana-Farber Cancer Institute.

This is a fully funded one-year fellowship. The fellow will receive research and career mentorship from clinician and AI faculty at Harvard-affiliated institutions, gain exposure to multimodal AI and LLM interpretability research, and have opportunities to publish in high-impact journals and present at national and international meetings. The experience is intended to provide a foundation for future residency, fellowship, physician-scientist, or industry careers in clinical AI.

Requirements

• Submit a CV, a cover letter describing research interests and goals, and a full list of publications by email to the hiring manager; applications submitted only through Workday may not be forwarded or reviewed.

Nice To Haves

• An MD or equivalent degree and/or experience in clinical medical practice, such as internship, residency, or fellowship training, is highly desirable.
• Medical students with appropriate clinical experience and candidates with other educational backgrounds and clinical experience are encouraged to apply.

Responsibilities

• Collaborate with computer scientists and clinicians on research into methods for detecting false outputs from multimodal AI models.
• Assist in designing, coordinating, and carrying out domain-expert clinician annotations and assessments of mechanistic interpretability methods in clinical medicine.
• Assist with analyzing clinician annotations and assessments.

Benefits

• Research and career mentorship from clinician and AI faculty at Harvard-affiliated institutions.
• Opportunities to publish in high-impact journals and present at national and international meetings.
• Fully funded one-year fellowship.

What This Job Offers

• Job Type: Full-time
• Career Level: Intern
• Education Level: No Education Listed