# Deconstructing Spatial Complexity: Hierarchical Decomposition for LLM Spatial Reasoning

arXiv: https://arxiv.org/abs/2605.28144 · HTML: https://mlsanity.com/paper/2605.28144/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning.
- Builds on 7 of its 41 references that are in the map.
- First posted on arXiv 27 May 2026.

## Placed in

- LLM reasoning / [Spatial and embodied reasoning](https://mlsanity.com/lr/spatial-and-embodied-reasoning/)

## Builds on (most cited first, 7 of 7)

- [A Survey of Large Language Models](https://mlsanity.com/paper/2303.18223/) (2023)
- [Understanding R1-Zero-Like Training: A Critical Perspective](https://mlsanity.com/paper/2503.20783/) (2025)
- [The Internal State of an LLM Knows When It's Lying](https://mlsanity.com/paper/2304.13734/) (2023)
- [Plan-and-Act: Improving Planning of Agents for Long-Horizon Tasks](https://mlsanity.com/paper/2503.09572/) (2025)
- [Self-Reflection in LLM Agents: Effects on Problem-Solving Performance](https://mlsanity.com/paper/2405.06682/) (2024)
- [AlphaMaze: Enhancing Large Language Models' Spatial Intelligence via GRPO](https://mlsanity.com/paper/2502.14669/) (2025)
- [Evolutionary Large Language Model for Automated Feature Transformation](https://mlsanity.com/paper/2405.16203/) (2024)

## Abstract (arXiv)

LLMs have shown remarkable proficiency in general language understanding and reasoning. However, they consistently underperform in spatial reasoning that severely limits their application, particularly in embodied intelligence. Inspired by the success of hierarchical reinforcement learning, this paper introduces a novel method for hierarchical task decomposition in LLM spatial reasoning. Our approach guides LLMs to decompose complex tasks into manageable sub-tasks by identifying key intermediate states and generating simplified sub-environments. However, we identify that LLMs often fail to derive optimal intermediate states due to their insufficient spatial prior, leading to sub-optimal task decomposition. To address this limitation and enhance its planning capability, we propose the MCTS-Guided Group Relative Policy Optimization (M-GRPO), where we reformulate the UCT formula by incorporating the LLM's prior predictive probabilities alongside its epistemic uncertainty. Furthermore, we implement a more fine-grained advantage function, enabling the model to learn optimal path planning. Experimental results demonstrate that our method substantially improves LLM performance on spatial tasks, including navigation, planning, and strategic games, achieving state-of-the-art results. This work paves the way for LLMs in real-world applications.