Initialization of Large Language Models via Reparameterization to Mitigate Loss Spikes Kosuke Nishida author Kyosuke Nishida author Kuniko Saito author 2024-11 text Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing Yaser Al-Onaizan editor Mohit Bansal editor Yun-Nung Chen editor Association for Computational Linguistics Miami, Florida, USA conference publication nishida-etal-2024-initialization 10.18653/v1/2024.emnlp-main.1264 https://aclanthology.org/2024.emnlp-main.1264/ 2024-11 22699 22714