Approaching the Source of Symbol Grounding with Confluent Reductions of Abstract Meaning Representation Directed Graphs

2025-08-14

AI Summary

This study addresses the symbol grounding problem by proposing a circuit-space-preserving graph reduction method for Abstract Meaning Representation (AMR) directed acyclic graphs (DAGs). To bridge the gap between symbolic AMR nodes and perceptual experience, we inject semantic knowledge from authoritative digital dictionaries into node representations and integrate contextualized embeddings from large-scale pretrained language models. We then design a flow-based reduction algorithm constrained by graph structural invariants, ensuring preservation of critical semantic paths and topological properties during compression. Experimental results demonstrate that the reduced AMR graphs significantly improve semantic interpretability and cross-instance fidelity. Crucially, this work establishes the first formally verifiable graph-structural analysis framework for symbol grounding—enabling rigorous, mathematically grounded evaluation of semantic grounding in AMR representations.

Technology Category

Machine Learning: Graph-based Machine LearningKnowledge Representation and Reasoning: ArgumentationCognitive Modeling & Cognitive Systems: Symbolic Representations

Application Category

Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphsSemantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semanticsSearch and Retrieval-Augmented AI: Web query analysis, representation and understanding

Abstract

Abstract meaning representation (AMR) is a semantic formalism used to represent the meaning of sentences as directed acyclic graphs. In this paper, we describe how real digital dictionaries can be embedded into AMR directed graphs (digraphs), using state-of-the-art pre-trained large language models. Then, we reduce those graphs in a confluent manner, i.e. with transformations that preserve their circuit space. Finally, the properties of these reduces digraphs are analyzed and discussed in relation to the symbol grounding problem.

Problem

• Embedding digital dictionaries into AMR directed graphs
• Confluent reduction of graphs preserving circuit space
• Analyzing reduced digraphs for symbol grounding problem

Innovation

• Embedding dictionaries into AMR digraphs using LLMs
• Confluent reduction preserving circuit space properties
• Analyzing reduced digraphs for symbol grounding