Towards a Neurosymbolic Reasoning System Grounded in Schematic Representations
This addresses the issue of unreliable logical reasoning in LLMs for AI applications, though it is incremental as it focuses on spatial primitives as a foundation.
The paper tackles the problem of LLMs being error-prone in logical reasoning by introducing Embodied-LM, a neurosymbolic system that grounds reasoning in schematic representations based on image schemas, resulting in enhanced interpretability and effective logical reasoning.
Despite significant progress in natural language understanding, Large Language Models (LLMs) remain error-prone when performing logical reasoning, often lacking the robust mental representations that enable human-like comprehension. We introduce a prototype neurosymbolic system, Embodied-LM, that grounds understanding and logical reasoning in schematic representations based on image schemas-recurring patterns derived from sensorimotor experience that structure human cognition. Our system operationalizes the spatial foundations of these cognitive structures using declarative spatial reasoning within Answer Set Programming. Through evaluation on logical deduction problems, we demonstrate that LLMs can be guided to interpret scenarios through embodied cognitive structures, that these structures can be formalized as executable programs, and that the resulting representations support effective logical reasoning with enhanced interpretability. While our current implementation focuses on spatial primitives, it establishes the computational foundation for incorporating more complex and dynamic representations.