AI Paper Atlas

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

FlowHijack: A Dynamics-Aware Backdoor Attack on Flow-Matching Vision-Language-Action Models

Xinyuan An, Tao Luo, Gengyun Peng, Yaobing Wang, Kui Ren, Dongxia Wang

CVPR 2026 · Adversarial robustness and security › Attacks on text and graphs · approach: LLM jailbreak and injection defense · 1 citations (OpenAlex)

Read the paper

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 FlowHijack dynamics-space attack instead of BadVLA with Object State triggerpi0 · LIBERO-GoalASR: 11.2 -> 100Improved
use FlowHijack Ours(PL) instead of BadVLA with Scene Semantic triggerpi0 · LIBERO-SpatialASR: 15.3 -> 100Improved
use Ours(PL) instead of BadVLA with White Pixel triggerpi0 · LIBERO-10SR (no trigger): 72.9 -> 82.0Improved
add the flow matching loss L_FM (ablation)FlowHijack · LIBERO suitesSR (no trigger): 0.0 -> 82.8Improved
add the backdoor hijacking loss L_BD (ablation)FlowHijack · LIBERO suitesASR: 0.0 -> 64.4Improved
add the dynamics mimicry regularizer L_mimic (ablation)FlowHijack · LIBERO-10ASR: 66.7 -> 64.4Mixed
add the dynamics mimicry regularizer L_mimic (ablation)FlowHijack · LIBERO-ObjectSR (no trigger): 95.3 -> 98.8Improved
enlarge injection window tau0 from 0.1 to 0.4 (ablation)Ours(IP) · LIBERO-GoalASR: 23.9 -> 100Improved
enlarge injection window tau0 from 0.4 to 0.8 (ablation)Ours(IP) · LIBERO-GoalSR (no trigger): 97.8 -> 96.6Hurt
tighten target position filter defense to 0.1 mOurs(PL) · LIBERO-GoalASR: 100 -> 17.8Hurt
tighten target position filter defense to 0.1 mOurs(IP) · LIBERO-GoalASR: 100 -> 82.2No clear effect
apply downstream clean LoRA fine-tuning for 10k stepsFlowHijack · LIBERO-GoalASR: 100 -> 67.7No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

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