Challenge: Existing certified robustness methods for certifying input-specific text perturbations have shown promise in certifyling UTPs, but masking only adversarial words can eliminate the attack.
Approach: They propose a method to certify a language model’s robustness against UTPs by using random smoothing.
Outcome: The proposed method achieves high certified accuracy under extensive masking and achieves state-of-the-art results in multiple settings.

Similar Papers

Close or Cloze? Assessing the Robustness of Large Language Models to Adversarial Perturbations via Word Recovery (2025.coling-main)

Copied to clipboard

Challenge: Existing models implicitly recover the original text, but it is unclear when they rely on context and when they implicitly do so.
Approach: They propose to use a dictionary to recover adversarial words by using a phonetic, typo, and visual attack to study word recovery performance.
Outcome: The proposed model outperforms open-source models on hateful, offensive, and toxic classification tasks.
PTP: Boosting Stability and Performance of Prompt Tuning with Perturbation-Based Regularizer (2023.emnlp-main)

Copied to clipboard

Challenge: Existing prompt tuning methods have training instability issues due to large variance of scores . existing prompt tuning algorithms have training stability issues due a slight change of input data .
Approach: They propose an algorithm that smooths the loss landscape of vanilla prompt tuning by perturbation-based regularizers.
Outcome: The proposed method improves the state-of-the-art prompt tuning methods by 1.94% and 2.34% on SuperGLUE and FewGLUE benchmarks.
Model-tuning Via Prompts Makes NLP Models Adversarially Robust (2023.emnlp-main)

Copied to clipboard

Challenge: Pre-trained models are typically adapted to downstream tasks by appending a randomly initialized multilayer perceptron to their topmost representation layer and fine-tuning the entire model on a downstream task.
Approach: They propose to append a multilayer perceptron to a CLS token and fine-tune the entire model on a downstream task.
Outcome: The proposed model-tuning via prompts outperforms adversarial training-based state-of-art defenses by 3.5% and improves against adversarials by 8% over standard methods.
PRP: Propagating Universal Perturbations to Attack Large Language Model Guard-Rails (2024.acl-long)

Copied to clipboard

Challenge: Recent work has shown that large language models are susceptible to automated jailbreak attacks that induce them to generate harmful content.
Approach: They propose a two-step attack strategy that leverages a universal adversarial prefix for the Guard Model and propagates this prefix to the response.
Outcome: The proposed attack strategy is successful against several open-source and closed-source implementations of Guard Models.
On Robustness of Prompt-based Semantic Parsing with Large Pre-trained Language Model: An Empirical Study on Codex (2023.eacl-main)

Copied to clipboard

Challenge: Existing techniques for parsing natural-language utterances are vulnerable to adversarial attacks, requiring large amounts of labelled data and expensive human annotation.
Approach: They propose to enhance the adversarial robustness of a prompt-based semantic parser based on a language model trained on code by constructing a set of demonstration examples.
Outcome: The proposed method can be enhanced without significant amounts of labelled data or expensive human annotations on in-domain semantic parsing data.
In and Out-of-Domain Text Adversarial Robustness via Label Smoothing (2023.acl-short)

Copied to clipboard

Challenge: Existing studies show that state-of-the-art NLP models are vulnerable to adversarial attacks . label smoothing has been proven effective in a variety of applications and modalities .
Approach: They propose to use label smoothing to improve adversarial robustness in pre-trained models against various popular attacks.
Outcome: The proposed method significantly improves adversarial robustness in pre-trained models against various popular attacks.
Certified Robustness to Word Substitution Attack with Differential Privacy (2021.naacl-main)

Copied to clipboard

Challenge: Recent studies have shown that adversarial examples can be easily fooled by DNNs, making the robustness and security of NLP models significantly important.
Approach: They propose a differential privacy-based algorithm to achieve certified robustness against word substitution at- tacks in text classification via differential privacy.
Outcome: The proposed model achieves higher accuracy and more than 30X efficiency improvement over existing defense algorithms.
Universal Adversarial Attacks with Natural Triggers for Text Classification (2021.naacl-main)

Copied to clipboard

Challenge: Recent work has demonstrated the vulnerability of modern text classifiers to universal adversarial attacks, which are input-agnostic sequences of words added to text processed by classifier.
Approach: They propose a gradient-based search that aims to maximize the downstream classifier’s prediction loss by using an adversarially regularized autoencoder to generate triggers and propose heuristics to spot such attacks.
Outcome: The proposed algorithms reduce model accuracy while being less identifiable than prior models as per automatic detection metrics and human-subject studies.
Calibrating LLM Confidence by Probing Perturbed Representation Stability (2025.emnlp-main)

Copied to clipboard

Challenge: Despite their impressive performance, large language models (LLMs) consistently struggle with confidence calibration.
Approach: They propose a method to analyze internal representational stability in large language models by applying adversarial perturbations to final hidden states and using a lightweight classifier to predict answer correctness.
Outcome: CCPS significantly outperforms existing methods on LLMs from 8B to 32B parameters in multiple-choice and open-ended formats.
MOCHA: Are Code Language Models Robust Against Multi-Turn Malicious Coding Prompts? (2025.findings-emnlp)

Copied to clipboard

Challenge: Recent advances in Large Language Models have significantly enhanced their code generation capabilities, but their robustness against adversarial misuse remains underexplored.
Approach: They introduce a code decomposition attack where a malicious coding task is broken down into subtasks across multiple conversational turns to evade safety filters.
Outcome: The proposed code decomposition attacks exploits multi-turn malicious coding prompts . the proposed model improves rejection rates while preserving coding ability .

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations