Natural Language Uncertainty in the Era of Large Language Models (LLMs)

Organizers

Geetanjali Bihani and Tatiana Ringenberg,
Department of Computer and Information Technology,
Purdue University

Abstract: Computing with words (CWW) as introduced by Zadeh, stated that it’s possible to have a system where the objects of computations are words, phrases and propositions drawn from natural language. With the emergence of Language Models in general, and Large Language Models (LLMs) in particular, the idea that mathematical solutions can be stated in natural language that could be understood by machines seems to be closer than ever. Yet, recent research has shown that computing with embedded representations (words, phrases, paragraphs) is successful in some areas but less successful in others (reasoning, hallucinations, trustworthiness, etc.).  This special session invites submissions that address successes and challenges in computing in the age of generative AI within natural language.