An Vo

I am a PhD student in Computer Science and Engineering at University of Michigan, advised by Rada Mihalcea. I also collaborate closely with Anh Totti Nguyen. Before joining Michigan, I worked as a Research Engineer at MBZUAI with Thamar Solorio and earned my MS from KAIST, where I received the Outstanding Thesis Award and was fully funded by the Hyundai CMK Global Scholarship.

My research focuses broadly on foundation models (LLMs/VLMs), with an emphasis on making them more trustworthy and explainable. My work has been used in industry by Google DeepMind, Alibaba Qwen, ByteDance, and Moonshot AI, and featured on Sky News. My current interests include:

1. Trustworthy & Explainable AI: understanding and diagnosing why models fail (e.g., bias, hallucination, robustness) through mechanistic interpretability or behavioral analysis.
2. Multimodal AI: understanding how vision and language interact in large models, particularly grounding, alignment, and cross-modal reasoning.