Increasing Computation Resolves Conflicts in Vision Language Models

TL;DR

This study shows that larger vision-language models exhibit human-like cognitive control, effectively resolving conflicts in visual tasks, with performance improving as model size increases, demonstrating emergent adaptive flexibility.

Contribution

It introduces a comprehensive benchmark for cognitive control in VLMs and reveals that larger models naturally develop conflict resolution abilities akin to human behavior.

Findings

Larger models resolve conflicts more effectively than smaller ones.

VLMs exhibit human-like conflict behavior, including dropping below chance on high-conflict trials.

Parameter count correlates with conflict resolution capacity.

Abstract

Cognitive control, the ability to coordinate competing information sources in pursuit of goals, is fundamental to intelligent behavior. We systematically investigate whether Vision Language Models (VLMs) exhibit cognitive control and how computational resources modulate conflict resolution. We construct a benchmark of 4,410 tasks across seven conflict paradigms (Stroop, Flanker, and five realistic variants) spanning multiple difficulty levels and visual complexities, testing 47 VLMs with rigorous experimental control. We find that VLMs exhibit robust congruency effects across all tasks, with larger models systematically resolving conflicts more effectively than smaller models. Critically, VLMs reproduce the fine-grained demand-resource relationship observed in human temporal dynamics: larger models drop below chance on incongruent high-conflict trials while smaller models fail to meaningfully engage and perform at chance, mirroring human behavior at short processing times and establishing parameter count as a proxy for conflict resolution capacity. These findings demonstrate that human-like cognitive control emerges from optimization dynamics in large-scale neural networks, suggesting that adaptive flexibility under conflict may naturally arise through scaling.