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
awesome-LLM-resources：GitHub 8K Star最全大语言模型资源库解析
githubWangRongsheng
Related Entities
Related Claims
UnverifiedLLaVA通过轻量级投影层将CLIP视觉编码器提取的图像特征映射到语言模型的嵌入空间，采用视觉编码器+投影层+语言模型的三段式结构77% similarUnverifiedTongying Lab has launched the Coin VLA vision-language-action model, focused on general embodied intelligence.69% similarVerifiedA VLM (vision-language model) can process both image and text inputs simultaneously, converting images into vector representations understandable by the model via a visual encoder, which are then fused with text information for reasoning.69% similarUnverifiedIPAdapter通过轻量级适配网络将参考图视觉特征编码为图像提示嵌入，与文本提示嵌入并行注入扩散模型的交叉注意力层，能捕捉难以用语言描述的视觉信息69% similarVerifiedMultimodal large models convert images into high-dimensional vector representations through a Vision Encoder, which are then fused with the text comprehension capabilities of the language model.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/31865API
curl https://kongchang.com/api/v1/knowledge/claims/31865MCP
get_claim(id=31865)