@misc{dl3-b-recursive-synthetic-support-1, title = {The Curse of Recursion: Training on Generated Data Makes Models Forget}, author = {Ilia Shumailov and Zakhar Shumaylov and Yiren Zhao and Yarin Gal and Nicolas Papernot and Ross Anderson}, year = {2023}, url = {https://arxiv.org/abs/2305.17493}, howpublished = {arXiv preprint}, archivePrefix = {arXiv}, eprint = {2305.17493}, note = {Cited author-manuscript edition; peer review is not inferred from arXiv} } @misc{dl3-b-recursive-synthetic-support-2, title = {Self-Consuming Generative Models Go MAD}, author = {Sina Alemohammad and Josue Casco-Rodriguez and Lorenzo Luzi and Ahmed Imtiaz Humayun and Hossein Babaei and Daniel LeJeune and Ali Siahkoohi and Richard G. Baraniuk}, year = {2023}, url = {https://arxiv.org/abs/2307.01850}, howpublished = {arXiv preprint}, archivePrefix = {arXiv}, eprint = {2307.01850}, note = {Cited author-manuscript edition; peer review is not inferred from arXiv} } @misc{dl3-b-recursive-synthetic-support-3, title = {Is Model Collapse Inevitable? Breaking the Curse of Recursion by Accumulating Real and Synthetic Data}, author = {Matthias Gerstgrasser and Rylan Schaeffer and Apratim Dey and et al.}, year = {2024}, url = {https://arxiv.org/abs/2404.01413}, howpublished = {arXiv preprint}, archivePrefix = {arXiv}, eprint = {2404.01413}, note = {Cited author-manuscript edition; peer review is not inferred from arXiv} } @misc{dl3-b-recursive-synthetic-support-4, title = {Self-Instruct: Aligning Language Models with Self-Generated Instructions}, author = {Yizhong Wang and Yeganeh Kordi and Swaroop Mishra and Alisa Liu and Noah A. Smith and Daniel Khashabi and Hannaneh Hajishirzi}, year = {2022}, url = {https://arxiv.org/abs/2212.10560}, howpublished = {arXiv preprint}, archivePrefix = {arXiv}, eprint = {2212.10560}, note = {Cited author-manuscript edition; peer review is not inferred from arXiv} } @misc{dl3-b-recursive-synthetic-support-5, title = {Self-training Improves Pre-training for Natural Language Understanding}, author = {Jingfei Du and Edouard Grave and Beliz Gunel and Vishrav Chaudhary and Onur Celebi and Michael Auli and Ves Stoyanov and Alexis Conneau}, year = {2020}, url = {https://arxiv.org/abs/2010.02194}, howpublished = {arXiv preprint}, archivePrefix = {arXiv}, eprint = {2010.02194}, note = {Cited author-manuscript edition; peer review is not inferred from arXiv} } @misc{dl3-b-recursive-synthetic-support-6, title = {On-Policy Distillation of Language Models: Learning from Self-Generated Mistakes}, author = {Rishabh Agarwal and Nino Vieillard and Yongchao Zhou and Piotr Stanczyk and Sabela Ramos and Matthieu Geist and Olivier Bachem}, year = {2023}, url = {https://arxiv.org/abs/2306.13649}, howpublished = {arXiv preprint}, archivePrefix = {arXiv}, eprint = {2306.13649}, note = {Cited author-manuscript edition; peer review is not inferred from arXiv} }