Current State

Current research focuses on understanding and mitigating the limitations of various AI models, particularly large language models (LLMs). Key areas include addressing biases in LLMs and developing effective "guardrails" to ensure safe and responsible deployment, as well as exploring the use of synthetic data to overcome data limitations in training and evaluating these models across diverse applications like medical imaging and autonomous vehicles. This work is crucial for advancing AI's reliability and trustworthiness, impacting fields ranging from healthcare and transportation to environmental sustainability and online safety.