article · Nature · 2024
AI models collapse when trained on recursively generated data
Ilia Shumailov, Zakhar Shumaylov, Yiren Zhao, Nicolas Papernot, Ross Anderson, Yarin Gal
Cited by 465Open access
Abstract
Stable diffusion revolutionized image creation from descriptive text. GPT-2 (ref. <sup>1</sup>), GPT-3(.5) (ref. <sup>2</sup>) and GPT-4 (ref. <sup>3</sup>) demonstrated high performance across a variety of language tasks. ChatGPT introduced such language models to the public. It is now clear that generative artificial intelligence (AI) such as large language models (LLMs) is here to stay and will substantially change the ecosystem of online text and images. Here we consider what may happen to GPT-{n} once LLMs contribute much of the text found online. We find that indiscriminate use of model-generated content in training causes irreversible defects in the resulting models, in which tails of the original content distribution disappear. We refer to this effect as 'model collapse' and show that it can occur in LLMs as well as in variational autoencoders (VAEs) and Gaussian mixture models (GMMs). We build theoretical intuition behind the phenomenon and portray its ubiquity among all learned generative models. We demonstrate that it must be taken seriously if we are to sustain the benefits of training from large-scale data scraped from the web. Indeed, the value of data collected about genuine human interactions with systems will be increasingly valuable in the presence of LLM-generated content in data crawled from the Internet.
Cite this paper
Shumailov, I., Shumaylov, Z., Zhao, Y., Papernot, N., Anderson, R., & Gal, Y. (2024). AI models collapse when trained on recursively generated data. Nature, 631(8022), 755–759. https://doi.org/10.1038/s41586-024-07566-y
Shumailov, Ilia, et al. “AI models collapse when trained on recursively generated data.” Nature, vol. 631, no. 8022, 2024, pp. 755–59, https://doi.org/10.1038/s41586-024-07566-y.
Shumailov, Ilia, Zakhar Shumaylov, Yiren Zhao, Nicolas Papernot, Ross Anderson, and Yarin Gal. 2024. “AI models collapse when trained on recursively generated data.” Nature 631 (8022): 755–59. https://doi.org/10.1038/s41586-024-07566-y.
Shumailov, I. et al. (2024) ‘AI models collapse when trained on recursively generated data’, Nature, 631(8022), pp. 755–759. Available at: https://doi.org/10.1038/s41586-024-07566-y.
@article{shumailov2024models,
title = {{AI models collapse when trained on recursively generated data}},
author = {Shumailov, Ilia and Shumaylov, Zakhar and Zhao, Yiren and Papernot, Nicolas and Anderson, Ross and Gal, Yarin},
journal = {Nature},
volume = {631},
number = {8022},
pages = {755--759},
year = {2024},
doi = {10.1038/s41586-024-07566-y}
}Made from the index record. Check names, capitalisation and page numbers against the paper. Need another style? AI for Research has over 10,000.