@article{reinforcementlearningamplifiesemergentmi,
  title = {Reinforcement Learning Can Amplify Emergent Misalignment from Harmless Rewards},
  author = {Magnus Jørgenvåg and David Kaczér and Lasse Ruttert and Marvin Gülhan and Lucie Flek and Florian Mai},
  year = {2026},
  eprint = {2605.31328},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2605.31328},
}