Grounded Theory

Grounded theory, in the context of AI and related fields, focuses on developing models and methods that connect abstract representations to real-world data and experiences, enhancing their reliability and interpretability. Current research emphasizes improving the factuality and diversity of generated text, developing methods for assessing and enhancing model trustworthiness, and creating benchmarks to evaluate the grounding of reasoning processes in various modalities (visual, textual, temporal). This work is crucial for building more robust, reliable, and explainable AI systems with applications ranging from question answering and information retrieval to healthcare and legal technology.