AI RESEARCH
To Agree or To Be Right? The Grounding-Sycophancy Tradeoff in Medical Vision-Language Models
arXiv CS.AI
•
ArXi:2603.22623v1 Announce Type: cross Vision-language models (VLMs) adapted to the medical domain have shown strong performance on visual question answering benchmarks, yet their robustness against two critical failure modes, hallucination and sycophancy, remains poorly understood, particularly in combination.