Unverified50% confidenceFactExact time
向量检索的效果高度依赖于分块策略，常用方法包括固定长度分块、基于语义的分块以及递归分块
1
Sources
50%
Confidence
Long-term
Relevance
8/23/2026
First Seen
Sources
Related Entities
Related Claims
Unverified高质量harness通常采用混合检索策略，将语义相似度与符号精确匹配结合，再通过重排序模型对候选片段打分73% similarUnverifiedCommon techniques for improving model efficiency include quantization, pruning, and architecture search.71% similarUnverifiedUltra mode produces different quality levels of output through differentiated allocation of compute resources at inference time—including longer reasoning chains, more verification steps, or higher sampling parameters70% similarUnverifiedChunk切割策略和Reranking往往比模型本身更决定RAG最终答案质量69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/788878API
curl https://kongchang.com/api/v1/knowledge/claims/788878MCP
get_claim(id=788878)