@article{howbadistrainingonsyntheticdataa, title = {How Bad is Training on Synthetic Data? A Statistical Analysis of Language Model Collapse}, author = {Mohamed El Amine Seddik and Suei-Wen Chen and Soufiane Hayou and Pierre Youssef and Merouane Debbah}, year = {2024}, eprint = {2404.05090}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2404.05090v1}, }