Multi-Modal, Pre-Trained Models in Verifiable Sequential Decision-Making - Metacognitive AI

Explore a 27-minute conference talk by Ufuk Topcu on "Multi-Modal, Pre-Trained Models in Verifiable Sequential Decision-Making" presented at the November 2023 ARO Metacognitive AI workshop. Delve into the intersection of multi-modal approaches, pre-trained models, and verifiable sequential decision-making in the context of Metacognitive AI. Gain insights into cutting-edge research and methodologies that combine symbolic methods and deep learning techniques. Learn how these advanced concepts contribute to the development of artificial general intelligence (AGI) and their potential applications in real-world scenarios. Discover the latest algorithms and approaches discussed in this presentation, which is part of a comprehensive workshop on metacognitive prediction of AI behavior. Access this talk through the Neuro Symbolic Channel, which offers a wealth of tutorials, courses, and research findings in the exciting field of neuro-symbolic AI, drawing from content originally developed for Arizona State University's AI course.