Human Cognitive Benchmarks Reveal Foundational Visual Gaps in MLLMs
This reveals foundational visual gaps in MLLMs, challenging assumptions about their perceptual capabilities, which is significant for AI researchers and developers aiming for human-like AI.
The paper tackles the gap between human and machine visual cognition by introducing VisFactor, a benchmark based on cognitive psychology tests, and finds that state-of-the-art MLLMs score only 25.19 out of 100, failing on basic tasks like mental rotation and spatial reasoning.
Despite significant progress on popular multimodal benchmarks, state-of-the-art Multimodal Large Language Models (MLLMs) continue to struggle with basic visual reasoning tasks that are trivially solved by humans, such as recognizing spatial relationships. To systematically investigate this gap, we introduce VisFactor, a benchmark that digitizes 20 vision-centric subtests from a well-established cognitive psychology assessment. These subtests span four core domains of human visual cognition: (1) Visualization and Spatial Processing, (2) Perceptual and Closure, (3) Memory, and (4) Reasoning. We evaluate 20 frontier MLLMs from GPT, Gemini, Claude, LLaMA, Qwen, and SEED families. The best-performing model achieves a score of only 25.19 out of 100, with consistent failures on tasks such as mental rotation, spatial relation inference, and figure-ground discrimination, regardless of model size or prompting strategy. These findings suggest that current MLLM performance gains on high-level benchmarks do not reflect human-like low-level visual cognition, challenging the assumption that large-scale pretraining naturally induces gestalt-like perceptual capabilities. The dataset and evaluation toolkit are publicly available at: https://github.com/CUHK-ARISE/VisFactor.