User-centric conversational AI, especially multi-lingual, multi-modal (speech/vision) use-cases. Improving LLM reasoning on multi-lingual, multi-modal or domain-specific tasks. Analyzing LLM behavior like sycophancy, toxicity & enhancing user safety.
Biography
Amruta is a second-year PhD student working with Prof. Hari Sundaram and Prof. Dilek Hakkani-Tur. Her research interests are in multimodal and conversational AI, specifically reasoning. She also works with multi-agent systems. She completed her B.Tech. in Electrical Engineering and M.Tech. in Artificial Intelligence and Data Science from IIT Bombay. Her Master's thesis was on multilingual speech recognition and generation, and her Bachelor's thesis was on medical image processing.