AI RESEARCH
CXR-ContraBench: Benchmarking Negated-Option Attraction in Medical VLMs
arXiv CS.CV
•
ArXi:2605.05810v1 Announce Type: new When a chest X-ray shows consolidation but the question asks which finding is present, a medical vision-language model may answer "No consolidation." This is than an incorrect choice: it is a polarity reversal that emits a clinical statement contradicting the image. We study this failure as negated-option attraction, where a model is drawn to a negated answer option even when it conflicts with both the visual evidence and the question. We