@article{howtransformersimplementinductionheads, title = {How Transformers Get Rich: Approximation and Dynamics Analysis}, author = {Mingze Wang and Ruoxi Yu and Weinan E and Lei Wu}, year = {2024}, eprint = {2410.11474}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2410.11474v3}, }