# LEGO-Puzzles: How Good Are MLLMs at Multi-Step Spatial Reasoning? [Kexian Tang](https://scholar.google.com/citations?user=cXjomd8AAAAJ&hl=zh-CN&oi=ao)1,2\*, [Junyao Gao](https://jeoyal.github.io/home/)1,2\*, [Yanhong Zeng](https://zengyh1900.github.io)1†, [Haodong Duan](https://kennymckormick.github.io/)1†, [Yanan Sun](https://scholar.google.com/citations?user=6TA1oPkAAAAJ&hl=zh-CN&oi=ao)1, [Zhening Xing](https://scholar.google.com/citations?hl=zh-CN&user=sVYO0GYAAAAJ)1, [Wenran Liu](https://scholar.google.com/citations?hl=zh-CN&user=fwKOaD8AAAAJ)1, [Kaifeng Lyu](https://kaifeng.ac/cn/)3‡, [Kai Chen](https://chenkai.site/)1‡ 1Shanghai AI Laboratory 2Tongji University 3Tsinghua University \*Equal contribution. †Project Leads. ‡Corresponding Authors.

arXiv PDF data img/data

## 🎉 News - **\[2025/04/08\]** The benchmark and evaluation code have been released! Have fun 😃 . - **\[2025/03/25\]** The paper is released. ## 📖 Introduction In this work, we introduce **LEGO-Puzzles**, a scalable and systematic benchmark designed to evaluate Multi-step Spatial Reasoning in Multimodal Large Language Models (MLLMs). Inspired by how humans develop spatial cognition through construction, LEGO-Puzzles frames spatial understanding as a series of LEGO assembly tasks that challenge both visual perception and sequential reasoning. To comprehensively assess spatial reasoning capabilities, LEGO-Puzzles is structured into three core task categories: **Spatial Understanding**, **Single-Step Sequential Reasoning**, and **Multi-Step Sequential Reasoning**. Each task requires models to understand visual inputs, perform step-by-step logical deduction, and maintain spatial consistency across sequences. Furthermore, based on LEGO-Puzzzles, we design **image generation tasks** to investigate whether MLLMs can transfer their spatial understanding and reasoning abilities to image generation. We further introduce **LEGO-Puzzles-Lite**, a distilled subset tailored for human-model comparison, and a fine-grained evaluation suite named **Next-k-Step** to test reasoning scalability under increasing complexity. Despite recent advances in multimodal modeling, our experiments reveal that current state-of-the-art MLLMs—while impressive—fall significantly short of human-level spatial reasoning, especially in multi-step and generative tasks. **LEGO-Puzzles aims to establish a rigorous testbed for benchmarking spatial reasoning in MLLMs and to motivate the development of more spatially-aware multimodal systems.** ## 🔍 Dataset & Task Design LEGO-Puzzles consists of **1,100 curated samples across 11 task types**, evenly covering: - 🧩 **Spatial Understanding** (36.4%) - 🔁 **Single-Step Sequential Reasoning** (36.4%) - 🧠 **Multi-Step Sequential Reasoning** (27.3%) Each task is framed as a visual question-answering problem or a generation prompt grounded in realistic LEGO configurations, enabling precise and interpretable evaluation.
## 🧪 Main Evaluation Results We evaluate **20 cutting-edge MLLMs**, spanning both open-source and proprietary models. While GPT-4o and Gemini-2.0-Flash lead overall, their performance still trails behind human annotators, especially in tasks requiring 3D spatial alignment, rotation handling, and multi-step assembly tracking.
## 👤 Human vs Model Performance To highlight the human-model performance gap, we compare top MLLMs against human annotators on **LEGO-Puzzles-Lite** (220 samples). Humans consistently outperform MLLMs by a wide margin, reaffirming the challenges of spatial reasoning in current AI systems.
## 🎨 Image Generation Evaluation We design **5 LEGO-based image generation tasks** testing a model's ability to simulate spatial transformations. Models must generate intermediate assembly states based on instructions. Human evaluators assess the output across two axes: - 🎯 **Appearance Similarity** - 🎯 **Instruction Following** Only **GPT-4o** and **Gemini-2.0-Flash** demonstrate partial success, while open-source models generally fail to produce structurally valid or instruction-aligned images. We evaluate GPT-4o, Gemini-2.0-Flash, GPT-4o* (referring to the version released prior to March 6, 2025), Emu2, GILL, and Anole using a scoring scale from 0 to 3 for both ***Appearance*** and ***Instruction Following*** dimensions.
## 🧠 Multi-Step Reasoning with Next-k-Step We propose **Next-k-Step**, a fine-grained reasoning benchmark that challenges models to predict assembly states after *k* sequential steps. We analyze model performance under varying values of *k*, both with and without **Chain-of-Thought (CoT)** prompting. Results suggest CoT does not robustly enhance multi-step spatial reasoning.
## 🧷 Qualitative Samples A few representative examples from LEGO-Puzzles are shown below, illustrating the diversity and complexity of the benchmark.
## 🛠️ Quick Start We have fully integrated **LEGO-Puzzles** into [VLMEvalKit](https://github.com/open-compass/VLMEvalKit), a unified framework for benchmarking VLMs. You can easily evaluate your favorite multimodal models on LEGO-Puzzles with just a single command! ### Step 0. Installation ```bash git clone https://github.com/open-compass/VLMEvalKit.git cd VLMEvalKit pip install -e . ``` ### Step 1. Setup API Keys (Optional) If you want to evaluate API-based models like GPT-4o, Gemini-Pro-V, etc., or use a LLM judge, configure the required keys in a .env file or export them as environment variables: ```bash # Example .env (place it in VLMEvalKit root directory) OPENAI_API_KEY=your-openai-key GOOGLE_API_KEY=your-google-api-key # ...other optional keys ``` If no key is provided, VLMEvalKit defaults to exact-match scoring (only works for Yes/No or multiple-choice tasks). ### Step 2. Run Evaluation on LEGO-Puzzles You can now run LEGO-Puzzles by simply setting the dataset name to `LEGO`. **Inference + Evaluation** ```bash python run.py --data LEGO --model --verbose # Example: # python run.py --data LEGO --model idefics_80b_instruct --verbose ``` **Inference Only** ```bash python run.py --data LEGO --model --verbose --mode infer # Example: # python run.py --data LEGO --model idefics_80b_instruct --verbose --mode infer ``` **Multi-GPU Acceleration (Optional)** ```bash torchrun --nproc-per-node=4 run.py --data LEGO --model --verbose # Example: # torchrun --nproc-per-node=4 run.py --data LEGO --model idefics_80b_instruct --verbose ``` ## Citation If you find LEGO-Puzzles useful, please cite using this BibTeX: ```bibtex @article{tang2025lego, title={LEGO-Puzzles: How Good Are MLLMs at Multi-Step Spatial Reasoning?}, author={Tang, Kexian and Gao, Junyao and Zeng, Yanhong and Duan, Haodong and Sun, Yanan and Xing, Zhening and Liu, Wenran and Lyu, Kaifeng and Chen, Kai}, journal={arXiv preprint arXiv:2503.19990}, year={2025} } ```