Understanding the Limits of Vision Language Models Through the Lens of the Binding Problem
This addresses a critical limitation in VLMs for applications requiring robust visual reasoning, though it is incremental in applying existing cognitive theories to explain model failures.
The paper investigates why state-of-the-art vision language models (VLMs) excel at describing and generating complex images but fail on basic multi-object reasoning tasks like counting and localization, linking these failures to the binding problem from cognitive science.
Recent work has documented striking heterogeneity in the performance of state-of-the-art vision language models (VLMs), including both multimodal language models and text-to-image models. These models are able to describe and generate a diverse array of complex, naturalistic images, yet they exhibit surprising failures on basic multi-object reasoning tasks -- such as counting, localization, and simple forms of visual analogy -- that humans perform with near perfect accuracy. To better understand this puzzling pattern of successes and failures, we turn to theoretical accounts of the binding problem in cognitive science and neuroscience, a fundamental problem that arises when a shared set of representational resources must be used to represent distinct entities (e.g., to represent multiple objects in an image), necessitating the use of serial processing to avoid interference. We find that many of the puzzling failures of state-of-the-art VLMs can be explained as arising due to the binding problem, and that these failure modes are strikingly similar to the limitations exhibited by rapid, feedforward processing in the human brain.