AI Paper Atlas

Research map / Adversarial robustness and security / Backdoor and poisoning

Exposing Functional Fusion: A New Class of Strategic Backdoor in Dynamic Prompt Architectures

Zeyao Liu, Zhendong Zhao, Xiaojun Chen, Xin Zhao, Yuexin Xuan, Xiaoshuang Ji

CVPR 2026 · Adversarial robustness and security › Backdoor and poisoning · approach: Detecting poisoned samples · 0 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 dynamic context-aware prompt generator instead of static VPT prompts (ablation)VIPER · Average over 6 datasetsACC (%): 82.81 -> 87.77Improved
use dynamic prompts instead of static prompts (ablation)VIPER · Food101ACC (%): 76.56 -> 89.95Improved
keep the Core (0.27% of parameters) in addition to Periphery (ablation)VIPER · ImageNet100ASR (%): 1.52 -> 100.00Improved
keep only the Core, dropping the Periphery (ablation)VIPER · ImageNet100ACC (%): 91.44 -> 91.10No clear effect
perturbative fine-tuning with poisoned images and random labels for one epoch (ablation)VIPER · ImageNet100ACC (%): 91.44 -> 2.52Hurt
add VIPER VPG to frozen ViT-B/16Clean ViT · ImageNet100ACC (%): 86.16 -> 91.44Improved
increase prompt length from 8 to 16 (ablation)VIPER · ImageNet100 and DTDACC (%): ImageNet100 91.44 -> 90.50; DTD 75.23 -> 76.85Mixed
decrease prompt length from 8 to 4 (ablation)VIPER · UCF101ACC (%): 82.37 -> 80.61Hurt
inject VPG at all 12 layers instead of layers 3, 6, 9 (ablation)VIPER · ImageNet100ACC (%): 91.44 -> 91.92No clear effect
inject VPG at shallow and middle layers 1-8 instead of 3, 6, 9 (ablation)VIPER · ImageNet100ACC (%): 91.44 -> 92.44No clear effect
replace LoRA attack with VIPER under 90% pruningVIPER vs LoRA attack · ImageNet100 pruning defenseASR (%): 100 at 90% pruning; LoRA collapses after 60%Improved
train on ImageNet100 and transfer zero-shot to five unseen datasetsVIPER · Cross-dataset transferACC (%): 83.13 -> 81.43No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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