Do Vision-Language Models Have Internal World Models? Towards an Atomic Evaluation

TL;DR

This paper systematically evaluates the internal world modeling capabilities of large vision-language models, revealing significant limitations in perception and prediction abilities compared to human-level understanding.

Contribution

It introduces a novel two-stage framework and WM-ABench benchmark for atomic evaluation of VLMs' world modeling abilities across diverse environments.

Findings

Models perform near-random in motion trajectory tasks.

Models lack disentangled understanding of object properties.

Significant gaps exist between VLMs and human world modeling.

Abstract

Internal world models (WMs) enable agents to understand the world's state and predict transitions, serving as the basis for advanced deliberative reasoning. Recent large Vision-Language Models (VLMs), such as OpenAI o3, GPT-4o and Gemini, exhibit potential as general-purpose WMs. While the latest studies have evaluated and shown limitations in specific capabilities such as visual understanding, a systematic evaluation of VLMs' fundamental WM abilities remains absent. Drawing on comparative psychology and cognitive science, we propose a two-stage framework that assesses Perception (visual, spatial, temporal, quantitative, and motion) and Prediction (mechanistic simulation, transitive inference, compositional inference) to provide an atomic evaluation of VLMs as WMs. Guided by this framework, we introduce WM-ABench, a large-scale benchmark comprising 23 fine-grained evaluation dimensions across 6 diverse simulated environments with controlled counterfactual simulations. Through 660 experiments on 15 latest commercial and open-source VLMs, we find that these models exhibit striking limitations in basic world modeling abilities. For instance, almost all models perform at near-random accuracy when distinguishing motion trajectories. Additionally, they lack disentangled understanding -- e.g., some models tend to believe blue objects move faster than green ones. More rich results and analyses reveal significant gaps between VLMs and human-level world modeling.