[KongchangAI]
Unverified50% confidenceFactExact time

扩散模型将学习到的视觉概念编码在高维潜空间(Latent Space)中,当提示词包含两种风格时,模型会找到两个风格向量的加权叠加点生成融合图像

1
Sources
50%
Confidence
Long-term
Relevance
7/11/2026
First Seen

Sources

Related Claims

Cite This Claim

Stable URI
https://kongchang.com/claim/486514
API
curl https://kongchang.com/api/v1/knowledge/claims/486514
MCP
get_claim(id=486514)