My primary research interests encompass enhancing the generalizability and robustness of AI models. This involves identifying and addressing failure modes of AI models, whether they stem from data quality, architectural design, or training algorithms. Additionally, I am actively involved in research related to 3D computer vision and graph neural networks.