Saved in:
Bibliographic Details
Main Authors: Born, Frieda, Neuhäuser, Tom, Muttenthaler, Lukas, Roads, Brett D., Spitzer, Bernhard, Lampinen, Andrew K., Jones, Matt, Müller, Klaus-Robert, Mozer, Michael C.
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.13883
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • 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.