Papers & Research

OpenAI Finds Deep Double Descent Phenomenon in Deep Learning

Original title:Deep double descent

AI summary

OpenAI reports that double descent is common in CNNs, ResNets and transformers: as model size, data volume or training time increases, performance first improves, then worsens, then improves again. The effect can often be avoided through careful regularization; the authors say its cause is not yet fully understood and list it as an important research direction.

Key Takeaways

  • OpenAI reports that double descent is common in CNNs,ResNets and transformers: as model size,data…
  • The effect can often be avoided through careful regularization
  • the authors say its cause is not yet fully understood and list it as an important research…

Quick Facts

Published
Updated
Source
OpenAI:官网动态(RSS · 排除企业/客户案例)
Category
Papers & Research
AI score
60
Read the original Read the full article at the original source
Why it matters

原文指出模型规模、数据量或训练时间增大时性能先降后升的普遍现象,并把它列为值得进一步研究的方向。

Related entities

Source & copyright

This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.

Original article · AIHOT entry

How to Cite

When citing this article, keep the title, date, and permalink, and credit Orange News as the source.

Orange News: OpenAI Finds Deep Double Descent Phenomenon in Deep Learning. December 5, 2019. https://new.nat6.net/en/news/g3wsst0c4ve4trvbvw1qw9ooc/

View all →