# SHIELD: Defending Textual Neural Networks against Multiple Black-Box Adversarial Attacks with Stochastic Multi-Expert Patcher

arXiv: https://arxiv.org/abs/2011.08908 · HTML: https://mlsanity.com/paper/2011.08908/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 Nov 2020.

## Built on by (most cited first, 13 of 13)

- [Navigating the Risks: A Survey of Security, Privacy, and Ethics Threats in LLM-Based Agents](https://mlsanity.com/paper/2411.09523/) (2024)
- [Granular-ball computing: an efficient, robust, and interpretable adaptive multi-granularity representation and computation method](https://mlsanity.com/paper/2304.11171/) (2023)
- [HQA-Attack: Toward High Quality Black-Box Hard-Label Adversarial Attack on Text](https://mlsanity.com/paper/2402.01806/) (2024)
- [TextShield: Beyond Successfully Detecting Adversarial Sentences in Text Classification](https://mlsanity.com/paper/2302.02023/) (2023)
- [Don't Retrain, Just Rewrite: Countering Adversarial Perturbations by Rewriting Text](https://mlsanity.com/paper/2305.16444/) (2023)
- [Risks, Causes, and Mitigations of Widespread Deployments of Large Language Models (LLMs): A Survey](https://mlsanity.com/paper/2408.04643/) (2024)
- [R&R: Metric-guided Adversarial Sentence Generation](https://mlsanity.com/paper/2104.08453/) (2021)
- [Improving Question Answering Performance Using Knowledge Distillation and Active Learning](https://mlsanity.com/paper/2109.12662/) (2021)
- [ROSE: Robust Selective Fine-tuning for Pre-trained Language Models](https://mlsanity.com/paper/2210.09658/) (2022)
- [On the Security Vulnerabilities of Text-to-SQL Models](https://mlsanity.com/paper/2211.15363/) (2022)
- [Identifying Human Strategies for Generating Word-Level Adversarial Examples](https://mlsanity.com/paper/2210.11598/) (2022)
- [Alert-ME: An Explainability-Driven Defense Against Adversarial Examples in Transformer-Based Text Classification](https://mlsanity.com/paper/2307.01225/) (2023)
- [SEP-Attack: A Simple and Effective Paradigm for Transfer-Based Textual Adversarial Attack](https://mlsanity.com/paper/2605.24958/) (2026)

## Abstract (arXiv)

Even though several methods have proposed to defend textual neural network (NN) models against black-box adversarial attacks, they often defend against a specific text perturbation strategy and/or require re-training the models from scratch. This leads to a lack of generalization in practice and redundant computation. In particular, the state-of-the-art transformer models (e.g., BERT, RoBERTa) require great time and computation resources. By borrowing an idea from software engineering, in order to address these limitations, we propose a novel algorithm, SHIELD, which modifies and re-trains only the last layer of a textual NN, and thus it "patches" and "transforms" the NN into a stochastic weighted ensemble of multi-expert prediction heads. Considering that most of current black-box attacks rely on iterative search mechanisms to optimize their adversarial perturbations, SHIELD confuses the attackers by automatically utilizing different weighted ensembles of predictors depending on the input. In other words, SHIELD breaks a fundamental assumption of the attack, which is a victim NN model remains constant during an attack. By conducting comprehensive experiments, we demonstrate that all of CNN, RNN, BERT, and RoBERTa-based textual NNs, once patched by SHIELD, exhibit a relative enhancement of 15%--70% in accuracy on average against 14 different black-box attacks, outperforming 6 defensive baselines across 3 public datasets. All codes are to be released.