AI Paper Atlas

Research map / Adversarial robustness and security / Attacks on text and graphs

VL-Trojan: Multimodal Instruction Backdoor Attacks against Autoregressive Visual Language Models

Jiawei Liang, Siyuan Liang, Aishan Liu, Xiaochun Cao

IJCV 2025 · Adversarial robustness and security › Attacks on text and graphs · approach: Vision-language model robustness · 77 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use VL-Trojan multimodal trigger vs Badnet (in-task)OpenFlamingo-3B · COCOASR (%): 37.30 -> 99.82Improved
use VL-Trojan trigger vs Badnet (spot-the-difference instruction data)OpenFlamingo-3B · COCOASR (%): 26.80 -> 99.98Improved
add L2 (cluster poisoned features) to L1 image trigger (ablation)VL-Trojan image trigger · COCOASR (%): 88.96 -> 96.68Improved
add L1 (separate poisoned from clean features) to L2 (ablation)VL-Trojan image trigger · COCOASR (%): 16.64 -> 96.68Improved
add text trigger to image trigger at 0.25% poisoning (ablation)VL-Trojan · COCOASR (%): 96.68 -> 97.86Improved
text trigger alone at 0.25% poisoning (ablation)VL-Trojan text trigger · COCOASR (%): 0.96No clear effect
add text trigger with surrogate ViT-B/16 (black-box) (ablation)VL-Trojan · COCOASR (%): 1.10 -> 97.72Improved
add text trigger with surrogate ViT-H/14 (black-box) (ablation)VL-Trojan · COCOASR (%): 0.68 -> 99.64Improved
add text trigger with surrogate RN50 (black-box) (ablation)VL-Trojan · COCOASR (%): 64.62 -> 99.78Improved
increase poisoning rate (ablation)VL-Trojan · COCOclean accuracyHurt
lower poisoning rate to 0.1% (23 samples) (ablation)VL-Trojan · COCOASR (%): exceeds 80%No clear effect
add more clean in-context examples (0 to 4 shot)conventional backdoors · COCOASR (%)Hurt

Later papers that compared with this method

3 later accepted papers report VL-Trojan as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy