@misc{indiciae38477d61b34b,
  title = {Noise Injection Systemically Degrades Large Language Model Safety Guardrails},
  author = {Prithviraj Singh Shahani and Kaveh Eskandari Miandoab and Matthias Scheutz},
  year = {2025},
  url = {https://arxiv.org/abs/2505.13500},
  note = {Source identifier: 2505.13500}
}