@article{simplemixfrustratinglysimplemixingofoff, title = {SIMPLEMIX: Frustratingly Simple Mixing of Off- and On-policy Data in Language Model Preference Learning}, author = {Tianjian Li and Daniel Khashabi}, year = {2025}, eprint = {2505.02363}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2505.02363v1}, }