[KongchangAI]
Partially Verified65% confidenceFactTime unknown

LLaVA用一个简单的线性层连接视觉编码器和语言模型,证明了视觉指令微调的有效性

3
Sources
65%
Confidence
Long-term
Relevance
6/1/2026
First Seen

Sources

Related Entities

Related Claims

Cite This Claim

Stable URI
https://kongchang.com/claim/31865
API
curl https://kongchang.com/api/v1/knowledge/claims/31865
MCP
get_claim(id=31865)