From Symbolic to Geometric: Enabling Spatial Reasoning in Large Language Models
2026-06-03 • Machine Learning
Machine LearningArtificial Intelligence
AI summaryⓘ
The authors explain that while large language models (LLMs) can seem to understand space, they mostly do this by matching words rather than truly understanding shapes and positions. To fix this, they created the Spatial Language Model (SLM), which uses special spatial information like real geometric data, not just text. They also made new data and tests to help train and check SLM's ability to reason about space better than older models that only use text clues. Their work shows adding real geometric thinking helps LLMs understand space more reliably.
large language modelsspatial reasoninggeometric reasoningmultimodal modelsspatial representationsinstruction datasetbenchmark evaluationprompt engineeringtopologyrelative position
Authors
Chen Chu, Bita Azarijoo, Li Xiong, Khurram Shafique, Cyrus Shahabi
Abstract
Recent large language models (LLMs) often appear to exhibit spatial reasoning ability; however, this capability is largely \emph{symbolic}, arising from pattern matching over spatial language rather than true \emph{geometric} reasoning over space. Because LLMs operate on discrete tokens, they lack native support for continuous spatial representations, explicit geometric computation, and structured spatial operators. To address this limitation, we introduce the \emph{Spatial Language Model (SLM)}, the first multimodal LLM that treats location information as a first-class modality and enables geometric spatial reasoning within the model's inference process. SLM directly operates on learned spatial representations rather than textual descriptions of spatial relations. To support effective training, we construct a \emph{Spatial Instruction Dataset} that aligns spatial representations, atomic geometric operations, and natural language instructions. We further propose a new benchmark named \emph{SpatialEval}, which is designed to evaluate spatial reasoning across attributes, distance, topology, and relative-position tasks. Extensive experiments show that SLM significantly outperforms existing LLM-based approaches that rely on symbolic reasoning via prompt engineering or textual abstraction, demonstrating the benefits of integrating geometric spatial representations for robust spatial reasoning. Our instruction dataset, evaluation benchmark, model training codes, and models' checkpoints can be found at: \hyperlink{https://github.com/chuchen2017/SLM}{https://github.com/chuchen2017/SLM}.