Unverified50% confidenceFactExact time
Embedding-based retrieval converts files or code blocks into vectors and stores them in an index, taking the most similar blocks at query time, which can turn stuffing the whole repo in into taking a small block by relevance, compressing the cost of a single call to a constant level.
1
Sources
50%
Confidence
Long-term
Relevance
10/14/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIndexing is the process of converting each segment's text into a vector via Embedding, and then storing both the segment text and its corresponding vector in a vector database.73% similarUnverifiedVector search works by embedding text into a high-dimensional space and using similarity measures to find the segments closest to the query.70% similarUnverifiedReranking is typically implemented using cross-encoders, which concatenate the query and candidate documents before feeding them in together; it usually reranks only the Top-50 to Top-100 candidates and finally selects Top-3 to Top-5 to pass to the generative model.70% similarUnverified倒排索引是搜索引擎核心数据结构，建立从词到文档的映射关系，使关键词检索时间复杂度接近O(1)70% similarUnverifiedKeyword retrieval tokenizes the query, counts the number of matched terms in each chunk, and scores and ranks them proportionally; it is simple to implement, costs nothing extra, and produces interpretable results, but can only match literal overlaps.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1027854API
curl https://kongchang.com/api/v1/knowledge/claims/1027854MCP
get_claim(id=1027854)