CyberdeceptionAtlas · Question → testES

Library / Paper

Paper

A Sweet Rabbit Hole by DARCY: Using Honeypots to Detect Universal Trigger's Adversarial Attacks

According to the abstract, the authors propose DARCY, a defense framework adapting the honeypot concept from cybersecurity to detect Universal Trigger textual adversarial attacks. DARCY injects multiple trapdoors into a neural network model to bait and detect such attacks, reporting high detection rates in experiments across four public datasets.

HoneypotDecoyApplication

What we know

According to the abstract, the authors propose DARCY, a defense framework adapting the honeypot concept from cybersecurity to detect Universal Trigger textual adversarial attacks. DARCY injects multiple trapdoors into a neural network model to bait and detect such attacks, reporting high detection rates in experiments across four public datasets.

Evidence limits

This entry is based solely on the published title and abstract; the full text has not been reviewed and findings have not been independently verified.

Reviewed: author abstract. Full-text results were not analyzed.

Sources and provenance

  1. A Sweet Rabbit Hole by DARCY: Using Honeypots to Detect Universal Trigger's Adversarial Attacks
    abstract · 2026-10-04

Reviewed: 2026-10-04. This record may change when new evidence is found.

RIS · BibTeX