AI Paper Atlas

Research map / Adversarial robustness and security / Backdoor and poisoning

BeDKD: Backdoor Defense Based on Directional Mapping Module and Adversarial Knowledge Distillation

Zhengxian Wu, Juan Wen, Wanli Peng, Yinghan Zhou, Changtong Dou, Yiming Xue

AAAI 2026 · Adversarial robustness and security › Backdoor and poisoning · approach: Training-time backdoor defense · 2 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
add DMM-identified poisoned data to fine-tuning (FT+DMM) (ablation)FT · SST2 (SynBkd)attack success rate: 66.22 -> 22.97Improved
add DMM-identified poisoned data to KD (KD+DMM) (ablation)KD · SST2 (SynBkd)attack success rate: 94.60 -> 40.54Improved
use adversarial KD (trust+punish) instead of KD with DMM (ablation)KD+DMM · SST2 (SynBkd)attack success rate: 40.54 -> 2.48Improved
add Lsoft to DMM distillation loss (ablation)DMM · SST2 (AddSent)FAR: 76.13 -> 0.00Improved
add Lsoft to DMM distillation loss (ablation)DMM · SST2 (SynBkd)FAR: 66.67 -> 42.12Improved
increase clean samples per class from 320 to 640 (ablation)BeDKD · SST2 (StyBkd)attack success rate: 4.86 -> 10.09Hurt
increase clean samples per class from 80 to 320 (ablation)BeDKD · SST2 (StyBkd)clean accuracy: 87.84 -> 90.59Improved
BeDKD vs W2SDefenseBeDKD · OLIDattack success rate: 17.61 -> 1.27Improved
BeDKD vs W2SDefenseBeDKD · SST2attack success rate: 27.25 -> 1.79Improved
BeDKD vs W2SDefenseBeDKD · AGnewsattack success rate: 3.34 -> 0.81Improved
BeDKD vs no defense on clean accuracyBeDKD · SST2clean accuracy: 90.27 -> 90.60No clear effect
lower poisoning rate from 10% to 5% on OLIDBeDKD · OLIDattack success rate: 1.35% -> 1.28% (average)No clear effect

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