cognitive architectures for trustworthy agents
Trust via reasoning that integrates LLMs and symbolic AI.
Trust via efficient execution that blends flexibility and controllability.
Trust via systematic & thinking about the world, task, and human partners.