@article{wheniswarmstartingeffectiveforscalinglan, title = {When is Warmstarting Effective for Scaling Language Models?}, author = {Neeratyoy Mallik and Maciej Janowski and Johannes Hog and Herilalaina Rakotoarison and Josif Grabocka and Frank Hutter and Aaron Klein}, year = {2026}, eprint = {2605.13405}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2605.13405}, }