Ignore the KL Penalty! Boosting Exploration on Critical Tokens to Enhance RL Fine-Tuning

@misc{vassoyan2025ignoreklpenalty,
      title={Ignore the KL Penalty! Boosting Exploration on Critical Tokens to  Enhance RL Fine-Tuning}, 
      author={Jean Vassoyan and Roman Plaud and Nathanaël Beau},
      year={2025},
      eprint={2502.06533},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2502.06533}, 
}