Full Professor, AIBE
3 papers at NeurIPS 2025
A game theory-based mechanism drives LLMs from self-check to consensus for improved accuracy.
NOVA is an extreme OOD stress-test dataset of ∼900 multi-modal brain MRI scans (with 281 rare pathologies) for benchmarking VLMs on three clinical tasks: anomaly localization, captioning, and diagnostic reasoning.
