Context Sensitivity Improves Human-Machine Visual Alignment
Abstract
Modern machine learning models typically represent inputs as fixed points in a high-dimensional embedding space. While this approach has been proven powerful for a wide range of downstream tasks, it fundamentally differs from the way humans process information. Because humans are constantly adapting to their environment, they represent objects and their relationships in a highly context-sensitive manner. To address this gap, we propose a method for context-sensitive similarity computation from neural network embeddings, applied to modeling a triplet odd-one-out task with an anchor image serving as simultaneous context. Modeling context enables us to achieve up to a 15% improvement in odd-one-out accuracy over a context-insensitive model. We find that this improvement is consistent across both original and "human-aligned" vision foundation models.
Cite
@article{arxiv.2604.13883,
title = {Context Sensitivity Improves Human-Machine Visual Alignment},
author = {Frieda Born and Tom Neuhäuser and Lukas Muttenthaler and Brett D. Roads and Bernhard Spitzer and Andrew K. Lampinen and Matt Jones and Klaus-Robert Müller and Michael C. Mozer},
journal= {arXiv preprint arXiv:2604.13883},
year = {2026}
}