Unverified50% confidenceFactExact time
假设文档嵌入（HyDE）先让LLM根据问题生成假设性回答，再用该假设回答检索真实文档以提升语义匹配精度
1
Sources
50%
Confidence
Long-term
Relevance
7/20/2026
First Seen
Sources
Related Claims
UnverifiedMainstream LLM text watermarking schemes embed statistical signals by manipulating the token sampling probability distribution during model generation, rather than inserting special characters.70% similarVerifiedHyDE (Hypothetical Document Embeddings) is a query rewriting strategy in which a large model first generates a hypothetical answer document, which is then used for retrieval.66% similarUnverifiedLLM本质上是在做条件概率预测，擅长回答字面问题而非用户真正需求，这种现象被称为字面遵从（Literal Compliance）65% similarUnverified任何将LLM与真实数据或工具链接的系统都必须在架构设计阶段将输入默认不可信作为基本假设65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/573310API
curl https://kongchang.com/api/v1/knowledge/claims/573310MCP
get_claim(id=573310)