@misc{indiciae29930b332069,
  title = {Efficient Safety Alignment of Language Models via Latent Personality Traits},
  author = {Mohamed Amine Merzouk and Nolan Smyth and Damiano Fornasiere and Linh Le and David Williams-King and Adam Oberman},
  year = {2026},
  url = {https://arxiv.org/abs/2607.07918},
  note = {Source identifier: 2607.07918}
}