作者位次 #5/6ACL 2026Findings· Findings of ACL
B io VLM : Routing Prompts, Not Parameters, for Cross-Modality Generalization in Biomedical VLM s
Mainak Singha · Tanisha Gupta · Ankit Jha · Muhammad Haris Khan · Sayantani Ghosh · Biplab Banerjee
摘要
Pretrained biomedical vision–language models (VLMs) such as BioMedCLIP perform well on average but often degrade on challenging modalities where inter-class margins are small and acquisition-specific variations are pronounced, especially under few-shot supervision and when modality priors differ...