Amaya Dharmasiri

I am a PhD student in the Princeton Visual AI Lab, advised by Dr. Olga Russakovsky. My research interests lie at the intersection of computer vision and machine learning, focusing on group-robustness and learning fair representations. I am broadly interested in building models that can learn fair, robust, generalizable, and data-efficient representations across modalities.

Before joining Princeton, I was a research assistant at the Mohamed bin Zayed University of Artificial Intelligence (MBZUAI), working with Dr. Salman H. Khan (MBZUAI) and Dr. Sadeep Jayasumana (Google Research). I received my Bachelor's degree in Computer Science and Engineering from the University of Moratuwa, Sri Lanka, where I completed my undergraduate thesis on learning representations for 3D point clouds and semantic manipulation of 3D point cloud objects, advised by Dr. Ranga Rodrigo and Dr. Kanchana Thilakarathna.

I am currently excited about a range of problems at the intersection of vision, language, and representation learning, including:

• Fairness and addressing biases in large-scale multi-modal models and data
• Distribution shift challenges of data accumulation in scaling up, and data selection in scaling down
• Scaling vision-language models while mitigating shortcut learning and bias
• Analysis of dataset-level and example-level mechanisms that lead to out-of-distribution generalization

Email  /  CV  /  Google Scholar  /  Twitter  /  Github