Towards a Neurosymbolic Reasoning System Grounded in Schematic Representations

2025-09-03

AI Summary

Large language models (LLMs) exhibit low accuracy and poor interpretability in logical reasoning due to their lack of embodied cognitive foundations. Method: We propose Embodied-LM, a neuro-symbolic language model integrating image schema theory—grounded in sensorimotor experience—with LLMs, using spatial structures as cognitive primitives to map natural language to executable spatial representations. Our approach synergistically combines LLM-based semantic understanding, image schema-based representation, and Answer Set Programming (ASP) for declarative spatial reasoning, enabling analogical inference and formal program generation. Contribution/Results: Experiments demonstrate substantial improvements in both accuracy and process transparency on logical deduction tasks. Embodied-LM advances the development of cognitively plausible architectures capable of human-level dynamic spatial representation and reasoning.

Technology Category

• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Intelligent Robots: Embodied AI
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

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.

Problem

• Addresses LLMs' logical reasoning errors and lack of robust mental representations
• Grounds understanding in schematic representations from sensorimotor experience
• Operationalizes spatial cognitive structures using declarative spatial reasoning

Innovation

• Neurosymbolic system grounds reasoning in schematic representations
• Uses spatial reasoning with Answer Set Programming
• Formalizes cognitive structures as executable programs