ResearchResearch paperLarge Language Models · Multimodal Models1 source · Oct 20, 2026

Gated Prototype Transfer for Zero-Support Medical Vision Language Model

Realistic few-shot adaptation of medical vision-language models (VLMs) can leave low-prevalence classes without any labeled support.

We in- troduce Gated Knowledge-Guided Prototype Transfer (GProtoT), a pre-repulsion module that transfers supported-class visual residuals to missing classes using VLM similarity and an ontology-based clinical distance.
A validation-free evidence gate scales transfer based on semantic relevance and residual agreement, thereby limiting unsupported extrapolation.
Relative to SS-Text+, gated ProtoT improved missing-class recall by 5.23 points, average class accuracy by 1.37 points, and the missing/supported recall harmonic mean by 1.83 points.
GProtoT is based on SS- Text+ and requires no gradients, validation data, or additional model inference, of- fering a low-compute path toward accountable adaptation under incomplete clini- cal supervision.

[1]Gated Prototype Transfer for Zero-Support Medical Vision Language Model
arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 20, 03:16 PM
Realistic few-shot adaptation of medical vision-language models (VLMs) can leave low-prevalence classes without any labeled support.
We in- troduce Gated Knowledge-Guided Prototype Transfer (GProtoT), a pre-repulsion module that transfers supported-class visual residuals to missing classes using VLM similarity and an ontology-based clinical distance.

Extractive summary: sentences quoted from the sources.

Before this