Model collapse
id:
model-collapse-169-11338290
title:
Model collapse
text:
Model collapse refers to a phenomenon where machine learning models gradually degrade due to errors coming from uncurated training on synthetic data, meaning the outputs of another model including prior versions of itself. Shumailov et al. coined the term and described two specific stages to the degradation: early model collapse and late model collapse. In early model collapse the model begins losing information about the tails of the distribution – mostly affecting minority data. Later work hig
brand slug:
wiki
category slug:
encyclopedia
description:
Degradation of AI models trained on synthetic data
original url:
https://en.wikipedia.org/wiki/Model_collapse
date created:
2024-03-06T14:05:14Z
date modified:
2024-08-31T12:02:09Z
main entity:
{"identifier":"Q123403903","url":"https://www.wikidata.org/entity/Q123403903"}
image:
fields total:
13
integrity:
15