Model collapse
Model collapse, also known as AI collapse, refers to the gradual degradation in the output of a generative artificial intelligence model trained on synthetic data, meaning the outputs of another model (including prior versions of itself).[1][2][3]
Repeating this process for generation after generation of models forms a so-called "autophagous" (self-consuming) loop. [4]
Theoretical and empirical analysis has demonstrated that, without enough fresh real data in each generation of an autophagous loop, future generative models are doomed to have their quality (precision) or diversity (recall) progressively decrease.
Recent work challenged the claim that frontier models are headed for model collapse, showing that model collapse does not occur under the realistic assumption that data accumulates over time.[5]
Impact on Language Models[edit]
In the context of Large Language Models, research found that training LLMs on predecessor-generated text—language models are trained on the synthetic data produced by previous models—causes a consistent decrease in the lexical, syntactic, and semantic diversity of the model outputs through successive iterations, notably remarkable for tasks demanding high levels of creativity.[6]
References[edit]
- ^ Mok, Aaron. "A disturbing AI phenomenon could completely upend the internet as we know it". Business Insider. Retrieved 2024-03-06.
- ^ Shumailov, Ilia; Shumaylov, Zakhar; Zhao, Yiren; Gal, Yarin; Papernot, Nicolas; Anderson, Ross (2023-05-31). "The Curse of Recursion: Training on Generated Data Makes Models Forget". arXiv:2305.17493 [cs.LG].
- ^ Ozsevim, Ilkhan (2023-06-20). "Research finds ChatGPT & Bard headed for 'Model Collapse'". Retrieved 2024-03-06.
- ^ Self-Consuming Generative Models Go MAD. The Twelfth International Conference on Learning Representations.
- ^ Gerstgrasser, Matthias; Schaeffer, Rylan; Dey, Apratim; Rafailov, Rafael; Sleight, Henry; Hughes, John; Korbak, Tomasz; Agrawal, Rajashree; Pai, Dhruv; Gromov, Andrey; Roberts, Daniel A.; Yang, Diyi; Donoho, David L.; Koyejo, Sanmi (2024-04-01). "Is Model Collapse Inevitable? Breaking the Curse of Recursion by Accumulating Real and Synthetic Data". arXiv:2404.01413 [cs.LG].
- ^ Guo, Yanzhu; Shang, Guokan; Vazirgiannis, Michalis; Clavel, Chloé (2024-04-16), The Curious Decline of Linguistic Diversity: Training Language Models on Synthetic Text, doi:10.48550/arXiv.2311.09807, retrieved 2024-05-06
 | This artificial intelligence-related article is a stub. You can help Wikipedia by expanding it. |