@article{spokesoptimizingfordiversepretrainingdat, title = {Spokes: Optimizing for Diverse Pretraining Data Selection}, author = {Clarence Lee and Yejin Choi and Luke Zettlemoyer and Pang Wei Koh and Hai Leong Chieu}, year = {2026}, eprint = {2606.15216}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2606.15216}, }