Sara Abdali

About

I am a Senior Researcher at Applied Sciences Group (ASG) working on a variety of AI/ML projects, including customization and distillation of Microsoft Turing and Microsoft Phi (opens in new tab) LLM/SLM for downstream tasks, multitask architectures, multimodal generative models, causal inferencing[1 (opens in new tab)], responsible GAI [2 (opens in new tab)3 (opens in new tab)4 (opens in new tab)] and more recently working/collaborating on multimodal conversational agents (opens in new tab) [5 (opens in new tab),6 (opens in new tab)], improving LLM’s reasoning [7 (opens in new tab)], and attention mechanism [8 (opens in new tab)]. Before joining Microsoft, I was a postdoctoral CIFellow (opens in new tab) at Georgia Tech (opens in new tab), selected for a fellowship program sponsored by CRA, CCC, and NSF to pursue my proposed research, “Adversarially robust multimodal misinformation detection” [9 (opens in new tab),10 (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, applying multilinear (tensor) algebra along with a variety of NLP and vision techniques [11 (opens in new tab),12 (opens in new tab),13 (opens in new tab),14 (opens in new tab)]. I also worked on Deepfake video detection [15 (opens in new tab)] and development of text augmentation techniques in few-shot settings [16 (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.