NLP Research Fellow
Core
Develop and test NLP models to detect bias, tone-policing, and equity violations in real-world healthcare review data.
Role type
NLP Research Fellow (Intern)
Builds
Bias detection systems and accountability tools for healthcare equity analytics
Domain
Healthcare + Natural Language Processing
Deliverable
production ML models
Required skills
Python, Hugging Face Transformers, spaCy, NLTK, zero-shot/few-shot classification, prompt engineering, model evaluation (sensitivity, specificity, fairness metrics), explainability tools (SHAP, LIME)
Preferred skills
Experience with MIMIC-III, Bias in Bios, or HealthReviewCorpus datasets, background in fairness or social impact
Technologies
Python, Hugging Face, spaCy, NLTK, SHAP, LIME
Responsibilities
Develop and test zero-shot and few-shot NLP models to flag bias and equity violations; Clean and label real-world review data from platforms like Google, RateMDs, Vitals, Healthgrades; Evaluate models for fairness metrics and false positives; Create explainability tools to make outputs auditable
Seniority
Intern
