Medical AI Evaluation Intern

**About the Role** We are looking for a Medical AI Evaluation Intern to benchmark large medical vision-language models on real-world clinical imaging data. You will work with a dataset of intraoral endoscopic images, run inference using state-of-the-art medical multimodal models, and systematically evaluate their performance on disease detection and classification tasks. This hands-on position is ideal for students or early-career researchers interested in the intersection of medical imaging, large language models, and clinical AI evaluation. You will gain direct exposure to the full evaluation pipeline — from data preparation and prompt engineering to metric design and analysis. **Responsibilities** - Curate, clean, and annotate a dataset of intraoral endoscopic images following clinical labeling guidelines. - Design and execute evaluation protocols for medical vision-language models, including ground-truth annotation, inference, and performance analysis. - Perform prompt engineering and workflow design to optimize model outputs for disease identification and classification. - Run systematic experiments to measure accuracy, recall, precision, and failure modes across different anatomical regions and pathological conditions. - Document findings in clear, reproducible reports and present results to the team. **Minimum Qualifications:** - Currently pursuing or recently completed a Bachelor's or Master's degree in Computer Science, Biomedical Engineering, Electrical Engineering, Data Science, or a related field. - Proficiency in Python. - Basic understanding of computer vision or medical image analysis concepts (classification, detection, segmentation, evaluation metrics). - Comfortable working with image datasets: loading, preprocessing, visualization, and annotation. - Strong attention to detail and ability to systematically document experimental procedures and results. **Preferred Qualifications:** - Experience with large language models or vision-language models (GPT, Gemini, LLaVA, Med-PaLM, or similar), including prompt engineering and API usage. - Exposure to medical imaging or healthcare AI projects. - Familiarity with at least one deep learning framework. - Experience with experiment tracking tools (Weights \& Biases, MLflow, or similar). - Ability to write lightweight Python scripts to automate data processing and evaluation workflows. - Basic understanding of statistical analysis and experiment design. ***- Mandarin Speaking is a plus*** **What You'll Gain** - Hands-on experience evaluating cutting-edge medical AI models on real clinical data. - Deep understanding of how large multimodal models perform in domain-specific medical tasks. - Mentorship from an interdisciplinary team spanning AI, software, and clinical applications. - Practical training in industry-grade medical AI product evaluation methodology. 职位类型：兼职, 实习生, 合同工 薪资： $20\.00至$25\.00（每小时 ） Work Location: Remote