From Impasse to Adaptation An Integrated LLM Supporting Meta Cognition in a Cognitive Architecture

Amitis Hamidi and Matthias Scheutz

We argue that successfully navigating the real world and its novelties, requires agents with meta-cognitive abilities that enable them to explore, adapt, and recover. We then propose a framework that shows such meta-cognitive capabilities can be realized in a cognitive architecture by integrating a large language model (LLM) as the engine for generating potential solutions when the agent encounters impasses during task performance that its regular planning and reasoning capabilities cannot solve.

@inproceedings{hamidisheutzaaaifs2026,
  title={From Impasse to Adaptation An Integrated LLM Supporting Meta Cognition in a Cognitive Architecture},
  author={Amitis Hamidi and Matthias Scheutz},
  year={2026},
  booktitle={Proceedings of the AAAI Fall Symposium Series},
  url={https://hrilab.tufts.edu/publications/hamidisheutzaaaifs2026.pdf}
}