Research

Artificial Metacognition is the ability of an AI system to reason about itself — a concept rooted in cognitive psychology that has re-emerged alongside large language models and neurosymbolic AI, driven by the need for more robust and secure AI systems. It goes beyond related areas like uncertainty quantification by unifying security, efficiency, explainability, and corrective action into one cognitively-inspired framework. The Symposium on Artificial Metacognition brings together researchers across computer science, cognitive psychology, and engineering to advance this field.

PyReason is a logic programming framework introduced in 2023 that combines ideas from generalized annotated programs and temporal logic programming to support modern AI workflows, including a Python interface and direct integration with machine learning models like PyTorch. It features CPU performance optimizations, reinforcement learning integration, and fuzzy logic support, and has been applied in fields like maritime operations, industrial automation, and medical triage. Learn more from the Leibniz Lab at Syracuse University, which created and maintains the project.