Navigate Complex Physical Worlds via Geometrically Constrained LLM
This addresses the problem of enabling LLMs to better understand and manipulate complex physical geometries for applications in robotics or simulation, though it appears incremental in combining existing techniques.
This study investigated whether large language models (LLMs) can reconstruct and construct physical worlds using only textual knowledge, and developed a workflow with geometric conventions, multi-layer graphs, and genetic algorithms to enhance their spatial understanding. The result demonstrated improved multi-step geometric inference capabilities for LLMs in spatial environments.
This study investigates the potential of Large Language Models (LLMs) for reconstructing and constructing the physical world solely based on textual knowledge. It explores the impact of model performance on spatial understanding abilities. To enhance the comprehension of geometric and spatial relationships in the complex physical world, the study introduces a set of geometric conventions and develops a workflow based on multi-layer graphs and multi-agent system frameworks. It examines how LLMs achieve multi-step and multi-objective geometric inference in a spatial environment using multi-layer graphs under unified geometric conventions. Additionally, the study employs a genetic algorithm, inspired by large-scale model knowledge, to solve geometric constraint problems. In summary, this work innovatively explores the feasibility of using text-based LLMs as physical world builders and designs a workflow to enhance their capabilities.