🛠️ Steven Gong

Search

SearchSearch

May 06, 2025, 1 min read

Contrastive Learning

Contrastive Loss

Resources

  • https://lilianweng.github.io/posts/2021-05-31-contrastive/#contrastive-loss

L=(1−y)⋅21​D2+y⋅21​{max(0,m−D)}2

Used for example in Siamese Networks.

Graph View

Backlinks

  • Contrastive Learning
  • Sigmoid Loss for Language Image Pre-Training (SigLIP)

Created with Quartz, © 2025

  • Blog
  • LinkedIn
  • Twitter
  • GitHub