RTD-Guard: A Black-Box Textual Adversarial Detection Framework via Replacement Token Detection

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhu, He, Li, Yanshu, Liu, Wen, Yang, Haitian
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914390444343296
author Zhu, He
Li, Yanshu
Liu, Wen
Yang, Haitian
author_facet Zhu, He
Li, Yanshu
Liu, Wen
Yang, Haitian
contents Textual adversarial attacks pose a serious security threat to Natural Language Processing (NLP) systems by introducing imperceptible perturbations that mislead deep learning models. While adversarial example detection offers a lightweight alternative to robust training, existing methods typically rely on prior knowledge of attacks, white-box access to the victim model, or numerous queries, which severely limits their practical deployment. This paper introduces RTD-Guard, a novel black-box framework for detecting textual adversarial examples. Our key insight is that word-substitution perturbations in adversarial attacks closely resemble the "replaced tokens" that a Replaced Token Detection (RTD) discriminator is pre-trained to identify. Leveraging this, RTD-Guard employs an off-the-shelf RTD discriminator-without fine-tuning-to localize suspicious tokens, masks them, and detects adversarial examples by observing the prediction confidence shift of the victim model before and after intervention. The entire process requires no adversarial data, model tuning, or internal model access, and uses only two black-box queries. Comprehensive experiments on multiple benchmark datasets demonstrate that RTD-Guard effectively detects adversarial texts generated by diverse state-of-the-art attack methods. It surpasses existing detection baselines across multiple metrics, offering a highly efficient, practical, and resource-light defense mechanism-particularly suited for real-world deployment in resource-constrained or privacy-sensitive environments.
format Preprint
id arxiv_https___arxiv_org_abs_2603_12582
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle RTD-Guard: A Black-Box Textual Adversarial Detection Framework via Replacement Token Detection
Zhu, He
Li, Yanshu
Liu, Wen
Yang, Haitian
Computation and Language
Cryptography and Security
Textual adversarial attacks pose a serious security threat to Natural Language Processing (NLP) systems by introducing imperceptible perturbations that mislead deep learning models. While adversarial example detection offers a lightweight alternative to robust training, existing methods typically rely on prior knowledge of attacks, white-box access to the victim model, or numerous queries, which severely limits their practical deployment. This paper introduces RTD-Guard, a novel black-box framework for detecting textual adversarial examples. Our key insight is that word-substitution perturbations in adversarial attacks closely resemble the "replaced tokens" that a Replaced Token Detection (RTD) discriminator is pre-trained to identify. Leveraging this, RTD-Guard employs an off-the-shelf RTD discriminator-without fine-tuning-to localize suspicious tokens, masks them, and detects adversarial examples by observing the prediction confidence shift of the victim model before and after intervention. The entire process requires no adversarial data, model tuning, or internal model access, and uses only two black-box queries. Comprehensive experiments on multiple benchmark datasets demonstrate that RTD-Guard effectively detects adversarial texts generated by diverse state-of-the-art attack methods. It surpasses existing detection baselines across multiple metrics, offering a highly efficient, practical, and resource-light defense mechanism-particularly suited for real-world deployment in resource-constrained or privacy-sensitive environments.
title RTD-Guard: A Black-Box Textual Adversarial Detection Framework via Replacement Token Detection
topic Computation and Language
Cryptography and Security
url https://arxiv.org/abs/2603.12582