AI RESEARCH
When Negation Is a Geometry Problem in Vision-Language Models
arXiv CS.CV
•
ArXi:2603.20554v1 Announce Type: new Joint Vision-Language Embedding models such as CLIP typically fail at understanding negation in text queries - for example, failing to distinguish "no" in the query: "a plain blue shirt with no logos". Prior work has largely addressed this limitation through data-centric approaches, fine-tuning CLIP on large-scale synthetic negation datasets. However, these efforts are commonly evaluated using retrieval-based metrics that cannot reliably reflect whether negation is actually understood.