IDEA:面向 GUI 智能體推理時後門檢測的意圖主導證據審計
IDEA: Intent-Dominance Evidence Audit for Inference-Time Backdoor Detection in GUI Agents
IDEA 提出一種面向開放權重 GUI 智能體的逐步推理時審計方法,用於檢測視覺觸發後門。方法比較指令和畫面區域對決策狀態的影響,並以乾淨步驟作對照,提供有限樣本的誤報率保證。Qwen2.5-VL-3B 在五類攻擊上的實驗顯示,IDEA 能在受控誤報率下檢測後門,並為每次警報定位觸發區域,推理開銷僅略有增加。
Published on Nov 18
Authors:
,
,
,
,
,
,
Abstract
A step-level inference-time audit detects visual-trigger backdoors in open-weight GUI agents by comparing instruction- and region-induced shifts in decision state against clean-step conformal baselines, localizing triggers with controlled false positives.
Organizations increasingly deploy downloaded open-weight GUI agents. A backdoored agent passes acceptance tests on clean data and misbehaves only when a visual trigger appears on screen, turning an unvetted checkpoint into a supply-chain risk. However, existing sample-level detectors target image classifiers and CLIP-style encoders, leaving step-level backdoor detection for GUI agents unaddressed. On clean steps, the instruction drives the action; when an unreferenced screen region hijacks the step, the instruction loses control over the decision state or that region dictates it instead. We term this signature intent--evidence dominance inversion and propose IDEA (Intent-Dominance Evidence Audit) to detect backdoor activation at inference time for each step without trigger knowledge or attack data. Specifically, Intent Counterfactual Probing (ICP) measures how far redirected instructions move the decision state, and Region Causal Audit (RCA) quantifies how far ablating a candidate region moves it. A gated conformal score compares both displacements against clean steps at the same distance from the clean manifold, providing a finite-sample false-positive guarantee. Extensive experiments on Qwen2.5-VL-3B across five attack families show that IDEA consistently detects backdoor activation under a controlled false-positive rate and outperforms state-of-the-art detection methods, while localizing the trigger region for every alarm and adding only modest inference overhead.
View arXiv page View PDF Add to collection
Models citing this paper 0
No model linking this paper
Cite arxiv.org/abs/2611.24102 in a model README.md to link it from this page.
Datasets citing this paper 0
No dataset linking this paper
Cite arxiv.org/abs/2611.24102 in a dataset README.md to link it from this page.
Spaces citing this paper 0
No Space linking this paper
Cite arxiv.org/abs/2611.24102 in a Space README.md to link it from this page.
Collections including this paper 0
No Collection including this paper
Add this paper to a collection to link it from this page.
來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co