Google, IBM & Meta Certificates — All 10,000+ Courses at 40% Off
One annual plan covers every course and certificate on Coursera. 40% off for a limited time.
Get Full Access
Explore a thought-provoking lecture from Georgia Institute of Technology's Santosh Vempala that delves into the fundamental questions of language model comprehension and cognitive boundaries. Learn about a proposed hierarchical framework for evaluating algorithmic understanding, applicable to both human and artificial intelligence. Examine empirical studies comparing human and GPT model comprehension levels, while discovering new insights into the phenomenon of AI hallucination. Understand the mathematical relationship between hallucination rates and misclassification in learning theory, including measurable lower bounds for factual hallucinations in language models. Drawing from research by Mirabel Reid and OpenAI's Adam T. Kalai, gain valuable perspectives on the intersection of cognitive science, linguistics, and neuroscience in the context of Large Language Models.
