Yin, Y. (2026). Use large language model to enhance reasoning of another large language model through reward updated GRPO. Scientific Reports. https://doi.org/10.1038/s41598-026-39296-8