Sara Abdali

About

I am a Senior Researcher at Applied Sciences Group (ASG) (opens in new tab) working on a variety of NLP and multimodal tasks including customization and distillation of Microsoft Turing (opens in new tab) and Microsoft Phi (opens in new tab) LLMs/SLMs for downstream tasks, multitask architectures, multimodal generative models, causal inferencing[1 (opens in new tab)], responsible GAI (opens in new tab)[2 (opens in new tab)3 (opens in new tab)4 (opens in new tab)] and more recently multimodal conversational agents (opens in new tab) [5 (opens in new tab)]. Before joining Microsoft, I was a postdoctoral CIFellow (opens in new tab) at Georgia Tech (opens in new tab), working on my NSF funded project, “Adversarially robust multimodal misinformation detection”[6 (opens in new tab),7 (opens in new tab)]. I earned my Ph.D. from the University of California, Riverside (UCR) (opens in new tab), where I received the Dean’s Distinguished Fellowship Award (opens in new tab), which fully funded my studies. During my Ph.D., I mainly worked on misinformation detection leveraging multilinear (tensor) algebra along with a variety of NLP and vision techniques [8 (opens in new tab),9 (opens in new tab),10 (opens in new tab),11 (opens in new tab)]. I also worked on Deepfake video detection [12 (opens in new tab)] and development of text augmentation techniques in few-shot settings [13 (opens in new tab)]. In fall 2020, I completed a research internship at Lenovo Research (opens in new tab), where I worked on deep learning-based image enhancement. In summer 2021, I finished another internship at Microsoft (opens in new tab) where I worked on development of cutting-edge NLP techniques.