@misc{indiciaebfa603140790,
  title = {Can Global XAI Methods Reveal Injected Behaviours in LLMs? SHAP vs Rule Extraction vs RuleSHAP},
  author = {Francesco Sovrano},
  year = {2026},
  doi = {10.1145/3770855.3818093},
  url = {https://arxiv.org/abs/2505.11189},
  note = {Source identifier: 2505.11189}
}