# Robustness of Large Language Models Against Adversarial Attacks

arXiv: https://arxiv.org/abs/2412.17011 · HTML: https://mlsanity.com/paper/2412.17011/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- Builds on 9 of its 18 references that are in the map.
- First posted on arXiv 22 Dec 2024.

## Builds on (most cited first, 9 of 9)

- [A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks](https://mlsanity.com/paper/1610.02136/) (2016)
- [Character-level Convolutional Networks for Text Classification](https://mlsanity.com/paper/1509.01626/) (2015)
- [How Can We Know What Language Models Know?](https://mlsanity.com/paper/1911.12543/) (2019)
- [Universal Adversarial Triggers for Attacking and Analyzing NLP](https://mlsanity.com/paper/1908.07125/) (2019)
- ["Do Anything Now": Characterizing and Evaluating In-The-Wild Jailbreak Prompts on Large Language Models](https://mlsanity.com/paper/2308.03825/) (2023)
- [On Evaluation of Adversarial Perturbations for Sequence-to-Sequence Models](https://mlsanity.com/paper/1903.06620/) (2019)
- [NEVLP: Noise-Robust Framework for Efficient Vision-Language Pre-training](https://mlsanity.com/paper/2409.09582/) (2024)
- [Twitter Sentiment Analysis of Covid Vacciness](https://mlsanity.com/paper/2311.04479/) (2023)
- [Financial Sentiment Analysis on News and Reports Using Large Language Models and FinBERT](https://mlsanity.com/paper/2410.01987/) (2024)

## Built on by (most cited first, 12 of 12)

- [Refusal Direction is Universal Across Safety-Aligned Languages](https://mlsanity.com/paper/2505.17306/) (2025)
- [CT-PatchTST: Channel-Time Patch Time-Series Transformer for Long-Term Renewable Energy Forecasting](https://mlsanity.com/paper/2501.08620/) (2025)
- [Enhancing Intent Understanding for Ambiguous prompt: A Human-Machine Co-Adaption Strategy](https://mlsanity.com/paper/2501.15167/) (2025)
- [Robustness in Large Language Models: A Survey of Mitigation Strategies and Evaluation Metrics](https://mlsanity.com/paper/2505.18658/) (2025)
- [MIRAGE: Multimodal Immersive Reasoning and Guided Exploration for Red-Team Jailbreak Attacks](https://mlsanity.com/paper/2503.19134/) (2025)
- [Research and Design on Intelligent Recognition of Unordered Targets for Robots Based on Reinforcement Learning](https://mlsanity.com/paper/2503.07340/) (2025)
- [Industrial Internet Robot Collaboration System and Edge Computing Optimization](https://mlsanity.com/paper/2504.02492/) (2025)
- [Efficient Temporal Consistency in Diffusion-Based Video Editing with Adaptor Modules: A Theoretical Framework](https://mlsanity.com/paper/2504.16016/) (2025)
- [TDRI: Two-Phase Dialogue Refinement and Co-Adaptation for Interactive Image Generation](https://mlsanity.com/paper/2503.17669/) (2025)
- [LLM in the Middle: A Systematic Review of Threats and Mitigations to Real-World LLM-based Systems](https://mlsanity.com/paper/2509.10682/) (2025)
- [Understanding the Ability of LLMs to Handle Character-Level Perturbation](https://mlsanity.com/paper/2510.14365/) (2025)
- [Active Imitation Learning for Thermal- and Kernel-Aware LFM Inference on 3D S-NUCA Many-Cores](https://mlsanity.com/paper/2604.11948/) (2026)

## Abstract (arXiv)

The increasing deployment of Large Language Models (LLMs) in various applications necessitates a rigorous evaluation of their robustness against adversarial attacks. In this paper, we present a comprehensive study on the robustness of GPT LLM family. We employ two distinct evaluation methods to assess their resilience. The first method introduce character-level text attack in input prompts, testing the models on three sentiment classification datasets: StanfordNLP/IMDB, Yelp Reviews, and SST-2. The second method involves using jailbreak prompts to challenge the safety mechanisms of the LLMs. Our experiments reveal significant variations in the robustness of these models, demonstrating their varying degrees of vulnerability to both character-level and semantic-level adversarial attacks. These findings underscore the necessity for improved adversarial training and enhanced safety mechanisms to bolster the robustness of LLMs.