@inproceedings{ul-abedin-etal-2025-arithmattack,
    title = "{A}rithm{A}ttack: Evaluating Robustness of {LLM}s to Noisy Context in Math Problem Solving",
    author = "Ul Abedin, Zain  and
      Qamar, Shahzeb  and
      Flek, Lucie  and
      Karimi, Akbar",
    editor = "Novikova, Jekaterina",
    booktitle = "Proceedings of the The First Workshop on LLM Security (LLMSEC)",
    month = aug,
    year = "2025",
    address = "Vienna, Austria",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/corrections-2025-08/2025.llmsec-1.5/",
    pages = "48--53",
    ISBN = "979-8-89176-279-4",
    abstract = "While Large Language Models (LLMs) have shown impressive capabilities in math problem-solving tasks, their robustness to noisy inputs is not well-studied. We propose ArithmAttack to examine how robust the LLMs are when they encounter noisy prompts that contain extra noise in the form of punctuation marks. While being easy to implement, ArithmAttack does not cause any information loss since words are not added or deleted from the context. We evaluate the robustness of eight LLMs, including LLama3, Mistral, Mathstral, and DeepSeek on noisy GSM8K and MultiArith datasets. Our experiments suggest that all the studied models show vulnerability to such noise, with more noise leading to poorer performances."
}