Saeid Asgari

About

My research is around developing interpretable and trustworthy (robust, fair, and reliable) machine learning systems that can handle domain shifts and spurious correlations without significant failures. Lately, my work has been focused on foundation models and leveraging language models for interpreting various models, including those related to vision. I am also actively working on addressing copyright and parroting (memorization) issues in large generative models.