# BaDExpert: Extracting Backdoor Functionality for Accurate Backdoor Input Detection

Source: [BaDExpert: Extracting Backdoor Functionality for Accurate Backdoor Input Detection](<https://www.sota2.com/research/paper/68fa70cf1e209131cea4d54a>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T01:50:52.513Z

Authors: Tinghao Xie, Xiangyu Qi, Ping He, Yiming Li, Jiachen T. Wang, Prateek Mittal

Published: 2023-08-23T21:47:06.000Z

## Abstract

We present a novel defense, against backdoor attacks on Deep Neural Networks (DNNs), wherein adversaries covertly implant malicious behaviors (backdoors) into DNNs. Our defense falls within the category of post-development defenses that operate independently of how the model was generated. The proposed defense is built upon a novel reverse engineering approach that can directly extract backdoor functionality of a given backdoored model to a backdoor expert model. The approach is straightforward -- finetuning the backdoored model over a small set of intentionally mislabeled clean samples, such that it unlearns the normal functionality while still preserving the backdoor functionality, and thus resulting in a model (dubbed a backdoor expert model) that can only recognize backdoor inputs. Based on the extracted backdoor expert model, we show the feasibility of devising highly accurate backdoor input detectors that filter out the backdoor inputs during model inference. Further augmented by an ensemble strategy with a finetuned auxiliary model, our defense, BaDExpert (Backdoor Input Detection with Backdoor Expert), effectively mitigates 17 SOTA backdoor attacks while minimally impacting clean utility. The effectiveness of BaDExpert has been verified on multiple datasets (CIFAR10, GTSRB and ImageNet) across various model architectures (ResNet, VGG, MobileNetV2 and Vision Transformer).

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2308.12439>)

[PDF](<https://arxiv.org/pdf/2308.12439.pdf>)
