Unverified50% confidenceOpinionExact time
EmbeddingGemma 2 very likely adopts a structure of a shared backbone plus modality adapters, rather than configuring a complete encoder for each modality.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
Unverified现代Sentence Embedding模型通常采用双塔（Bi-Encoder）结构，查询和文档分别独立编码为向量68% similarUnverified一些LLaVA变体尝试同时使用CLIP和DINOv2等自监督编码器实现语义与结构特征互补67% similarUnverifiedEmbeddingGemma 2 adopts Matryoshka Representation Learning, allowing dimensions of 256, 128, or fewer to still be used as standalone embeddings.64% similarUnverified主流Embedding模型（如BGE、text-embedding系列）通常基于BERT架构的编码器（Encoder-only），通过对比学习让语义相近的句子在向量空间中靠拢；而GPT类生成模型基于解码器（Decoder-only）64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1007011API
curl https://kongchang.com/api/v1/knowledge/claims/1007011MCP
get_claim(id=1007011)