π©π»βπ» I'm a CS PhD candidate at UIUC's Conversational AI Lab, advised by Prof. Dilek Hakkani-TΓΌr.
π¬ My research focuses on LLM alignment, evaluation, and safety β figuring out when, how, and why language models become sycophant, biased, or deceptive, and how to fix them.
π« Reach me at: rabbani8@illinois.edu
π Website Β· Google Scholar Β· LinkedIn
π£οΈ DialDefer β measuring LLM judgment shift based on who's talking, not what's being said
π¬ From Fact to Judgment β investigating task framing effects on LLM conviction (IWSDS'26)
My baby Blu π
