Saved in:
Bibliographic Details
Main Authors: Fang, Yunxuan, Zhang, Ziwei, Wang, Xinhe
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.09549
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909030788628480
author Fang, Yunxuan
Zhang, Ziwei
Wang, Xinhe
author_facet Fang, Yunxuan
Zhang, Ziwei
Wang, Xinhe
contents Adaptive prompting mechanisms have been proposed to enhance vision-language models by dynamically tailoring prompts to inputs. However, in frozen few-shot prompt learning with CLIP-style backbones, we systematically observe that adaptive gates and prompt-selection modules often collapse: they produce nearly constant outputs, contribute negligible gradient signals, and frequently fail to outperform fixed prompts. To further explore this issue, we present a systematic diagnostic study to uncover the underlying causes and conditions of adaptation failure. Through controlled experiments across datasets and multiple prompt learning architectures, we identify two recurring failure modes: gradient magnitude imbalance and gate degradation. Our findings invite a re-examination of indiscriminately adding architectural complexity in parameter-efficient learning and clarify when prompt-level adaptive gating is, and is not, effective in this regime.
format Preprint
id arxiv_https___arxiv_org_abs_2605_09549
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle When Adaptation Fails: A Gradient-Based Diagnosis of Collapsed Gating in Vision-Language Prompt Learning
Fang, Yunxuan
Zhang, Ziwei
Wang, Xinhe
Machine Learning
Adaptive prompting mechanisms have been proposed to enhance vision-language models by dynamically tailoring prompts to inputs. However, in frozen few-shot prompt learning with CLIP-style backbones, we systematically observe that adaptive gates and prompt-selection modules often collapse: they produce nearly constant outputs, contribute negligible gradient signals, and frequently fail to outperform fixed prompts. To further explore this issue, we present a systematic diagnostic study to uncover the underlying causes and conditions of adaptation failure. Through controlled experiments across datasets and multiple prompt learning architectures, we identify two recurring failure modes: gradient magnitude imbalance and gate degradation. Our findings invite a re-examination of indiscriminately adding architectural complexity in parameter-efficient learning and clarify when prompt-level adaptive gating is, and is not, effective in this regime.
title When Adaptation Fails: A Gradient-Based Diagnosis of Collapsed Gating in Vision-Language Prompt Learning
topic Machine Learning
url https://arxiv.org/abs/2605.09549