@misc{indiciaef6e74d38c9cf,
  title = {Optimus: A Robust Defense Framework for Mitigating Toxicity while Fine-Tuning Conversational AI},
  author = {Aravind Cheruvu and Shravya Kanchi and Sifat Muhammad Abdullah and Nicholas Ka-Shing Kong and Daphne Yao and Murtuza Jadliwala and Bimal Viswanath},
  year = {2026},
  url = {https://arxiv.org/abs/2507.05660},
  note = {Source identifier: 2507.05660}
}