@article{balancingenhancementharmlessnessand,
  title = {Balancing Enhancement, Harmlessness, and General Capabilities: Enhancing Conversational LLMs with Direct RLHF},
  author = {Chen Zheng and Ke Sun and Hang Wu and Chenguang Xi and Xun Zhou},
  year = {2024},
  eprint = {2403.02513},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2403.02513v1},
}