@misc{indiciae521e60218062,
  title = {Deceptive Automated Interpretability: Language Models Coordinating to Fool Oversight Systems},
  author = {Simon Lermen and Mateusz Dziemian and Natalia Pérez-Campanero Antolín},
  year = {2025},
  url = {https://arxiv.org/abs/2504.07831},
  note = {Source identifier: 2504.07831}
}