跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.19868 · v1

HalluProbe: Exploiting VLM Hallucinations for Backdoor Detection

首次提交 2026-11-15 16:16
版本更新 2026-11-15 16:16
雷达观测 2026-11-17 12:06
arXiv ↗PDF ↗

摘要

Vision-Language Models (VLMs) are vulnerable to backdoor attacks introduced through data poisoning, making effective backdoor detection a critical security challenge. In this paper, we uncover a novel connection between VLM hallucinations and backdoor behaviors and exploit this connection for backdoor detection. Our key insight is that poisoned training data induces strong language priors toward the backdoor target token, while such priors are also a major source of hallucinations. This suggests that enhancing hallucinations can expose hidden backdoor-related priors. Specifically, backdoored VLMs tend to hallucinate backdoor target tokens more frequently across diverse inputs, while the hallucinated tokens of clean VLMs are largely random and exhibit no stable target-specific bias. Based on this observation, we propose HalluProbe, which enhances hallucinations during output decoding and measures the distributional consistency of hallucinated tokens to identify backdoored VLMs. We evaluate HalluProbe against six representative VLM backdoor attacks across three models. Experimental results demonstrate that HalluProbe can effectively detect backdoored VLMs with an accuracy exceeding 90%, outperforming existing backdoor detection methods.

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

目前只收录了一个版本。收录下一版后,可以对照标题、作者、分类和摘要。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v1 · 2026-11-15 16:16 · 最新观测版本

HalluProbe: Exploiting VLM Hallucinations for Backdoor Detection

Vision-Language Models (VLMs) are vulnerable to backdoor attacks introduced through data poisoning, making effective backdoor detection a critical security challenge. In this paper, we uncover a novel connection between VLM hallucinations and backdoor behaviors and exploit this connection for backdoor detection. Our key insight is that poisoned training data induces strong language priors toward the backdoor target token, while such priors are also a major source of hallucinations. This suggests that enhancing hallucinations can expose hidden backdoor-related priors. Specifically, backdoored VLMs tend to hallucinate backdoor target tokens more frequently across diverse inputs, while the hallucinated tokens of clean VLMs are largely random and exhibit no stable target-specific bias. Based on this observation, we propose HalluProbe, which enhances hallucinations during output decoding and measures the distributional consistency of hallucinated tokens to identify backdoored VLMs. We evaluate HalluProbe against six representative VLM backdoor attacks across three models. Experimental results demonstrate that HalluProbe can effectively detect backdoored VLMs with an accuracy exceeding 90%, outperforming existing backdoor detection methods.

查看此版本解读 →查看此版本原文 ↗