EmbodiedCity/BasicSpatialAbility

[ACL'25 Main] Defining and Evaluating Visual Language Models’ Basic Spatial Abilities: A Perspective from Psychometrics

You can find the sample testing code on GitHub!

This dataset is a benchmark designed for evaluating Multimodal Large Language Models' Basic Spatial Abilities based on authentic Psychometric theories. It is structured specifically to support both Zero-shot and Few-shot evaluation protocols.

Split Name	Role	Description
test	Query Set	Contains the actual benchmark questions (images & queries) to be evaluated. ⚠️ Evaluation Only. Do not use for training or as few-shot examples.
validation	Support Set	Contains high-quality examples intended to be used as Few-shot Prompts (In-Context Learning). These samples should be prepended to the test queries to demonstrate the task to the model.

⚙️ Usage & Evaluation Protocol

You can load the dataset using the Hugging Face datasets library.

1. Zero-Shot Evaluation

Logic: Directly evaluate the model on the test split without any prior examples.

from datasets import load_dataset

# Load the evaluation queries
test_dataset = load_dataset("EmbodiedCity/BasicSpatialAbility", split="test")

for sample in test_dataset:
    image = sample['image']
    question = sample['question']
    # Model inference...

2. Few-Shot Evaluation

Logic: Use examples from the validation split as the context (demonstrations), followed by the query from the test split.

1. Load the validation split.
2. Format them into the prompt history.
3. Append the target question from the test split.

from datasets import load_dataset

# 1. Load the support set (demonstrations)
support_set = load_dataset("EmbodiedCity/BasicSpatialAbility", split="validation")

# 2. Load the query set (evaluation)
test_set = load_dataset("EmbodiedCity/BasicSpatialAbility", split="test")

# Pseudo-code for prompt construction
prompt_context = []
for ex in support_set:
    prompt_context.append(f"User: {ex['question']}\nAssistant: {ex['answer']}")

# 3. Evaluate on Test Set
for sample in test_set:
    # Combine context + current test question
    final_prompt = prompt_context + [f"User: {sample['question']}"]
    
    # Model inference...

🔬 Underlying Theory

The Theory of Multiple Intelligences underscores the hierarchical nature of cognitive capabilities. To advance Spatial Artificial Intelligence, we pioneer a psychometric framework defining five Basic Spatial Abilities (BSAs) in Visual Language Models (VLMs): Spatial Perception, Spatial Relation, Spatial Orientation, Mental Rotation, and Spatial Visualization. Benchmarking 13 mainstream VLMs through nine validated psychometric experiments reveals significant gaps versus humans, with three key findings: 1) VLMs mirror human hierarchies (strongest in 2D orientation, weakest in 3D rotation) with independent BSAs; 2) Many smaller models surpass larger counterparts, with Qwen leading and InternVL2 lagging; 3) Interventions like CoT and few-shot training show limits from architectural constraints, while ToT demonstrates the most effective enhancement. Identified barriers include weak geometry encoding and missing dynamic simulation. By linking Psychometrics to VLMs, we provide a comprehensive BSA evaluation benchmark, a methodological perspective for embodied AI development, and a cognitive science-informed roadmap for achieving human-like spatial intelligence.

Type	Definition	Tests
Spatial Perception	The ability to perceive horizontal and vertical orientations without interference from miscellaneous information.	SVT
Spatial Relation	The ability of recognizing relationships between parts of an entity.	NCIT DAT:SR R-Cube-SR
Spatial Orientation	The ability to navigate or enter a given spatial state.	MRMT
Mental Rotation	The ability to mentally rotate 3D objects.	MRT PSVT:R
Spatial Visualization	The ability to mentally manipulate and transform 2D and 3D objects.	SBST R-Cube-Vis

The Framework of Basic Spatial Abilities (Image sources are cited in the paper)

Citation

If you use this project in your research, please cite the following paper:

@inproceedings{xu-etal-2025-defining,
    title = "Defining and Evaluating Visual Language Models' Basic Spatial Abilities: A Perspective from Psychometrics",
    author = "Xu, Wenrui  and
      Lyu, Dalin  and
      Wang, Weihang  and
      Feng, Jie  and
      Gao, Chen  and
      Li, Yong",
    booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2025",
    address = "Vienna, Austria",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2025.acl-long.567/",
    doi = "10.18653/v1/2025.acl-long.567",
    pages = "11571--11590",
    ISBN = "979-8-89176-251-0"
}