@misc{indiciaefd12bea6add4,
  title = {Neural Chameleons: Language Models Can Learn to Hide Their Thoughts from Unseen Activation Monitors},
  author = {Max McGuinness and Alex Serrano and Luke Bailey and Scott Emmons},
  year = {2025},
  url = {https://arxiv.org/abs/2512.11949},
  note = {Source identifier: 2512.11949}
}