// ◉ Multimodal & Vision
CLIP (Contrastive Language–Image Pretraining)
Aligns images and text in one embedding space from 400M pairs.
CLIP 图文对比学习 2021 Foundational core
IN PLAIN WORDS
Image and text encoders are trained so matched pairs in a batch score high and mismatched ones low. It enables zero-shot classification and text-image search, and conditions text-to-image models.
Used in: Stable Diffusion text encoder, image search
TIMELINE · prerequisites → this → what builds on it
01Original paper
Learning Transferable Visual Models From Natural Language Supervision
Alec Radford, Jong Wook Kim, Chris Hallacy, et al. · 2021
arXiv ↗ ✓ Verified · arXiv API
02Where it sits
Learn the left column first; the right column builds on this one.
Left: prerequisites · right: builds on this · dashed: related. Full map →
03Learning resources
Official docs, explainers, courses and reference code. All links checked.
Reference code
Explainers
04Where the course teaches it
On learning paths:Generative Models & Frontier Research
For AI / data
- Verification: verified, checked 2026-10-21.
- JSON: /api/learn/algorithms.json (id:
clip)