معرفی
Dr. Andi Zhang is a Machine Learning Research Associate at the Centre for AI Fundamentals, University of Manchester, specializing in out-of-distribution detection and probabilistic modeling. His research develops methods for improving model robustness and uncertainty quantification in deep learning systems.
Research Contributions: Pioneered techniques for using fine-tuned LLMs as OOD detectors without architectural modifications. Developed novel approaches for semantic-aware adversarial example generation through probabilistic frameworks. Contributed fundamental critiques of misconceptions in OOD detection literature.
Technical Expertise: Bayesian optimization, neural architecture analysis, and generative modeling. Maintains active GitHub repositories documenting experimental frameworks for OOD detection and adversarial robustness.
