Explore a thought-provoking 20-minute talk by Eunice Yiu from UC Berkeley, delving into the fascinating world of KiVA (Kid-inspired Visual Analogies) and its application to large multimodal models. Gain insights into how this innovative approach draws inspiration from children's cognitive processes to enhance the capabilities of AI systems. Discover the intersection of artificial intelligence, psychology, and neuroscience as Yiu examines higher-level intelligence from multiple perspectives. Learn about the potential implications of KiVA for advancing our understanding of both human and machine cognition, and how it could shape the future of multimodal AI development.
KiVA - Kid-inspired Visual Analogies for Large Multimodal Models
Overview
Syllabus
KiVA: Kid-inspired Visual Analogies for Large Multimodal Models
Taught by
Simons Institute
Related Courses
-
Building Multimodal Search and RAG
-
Accelerate Your Learning with ChatGPT
5.0 -
How Grounded Are Ungrounded LLM Models
-
From Large Language Models to Large Multimodal Models - Stanford CS25 - Lecture 4
-
A Multimodal Automated Interpretability Agent - Understanding Lower-Level Intelligence
-
Consciousness and Large Language Models - Understanding Higher-Level Intelligence
