Unverified50% confidenceFactExact time
chatVia.ai uses cross-encoder rerankers (Cohere, Jina, Voyage, etc.) to rerank the fused candidates, and sets a similarity gate to discard weak matches.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/30/2026
First Seen
Valid until: 1/28/2027
Sources
Related Entities
Related Claims
UnverifiedFor each query, chatVia.ai runs in parallel a pgvector-based dense vector similarity search and an in-Postgres sparse BM25 keyword search, and merges the rankings using Reciprocal Rank Fusion.75% similarUnverifiedRAG的pipeline分为切块、向量化、混合检索、重排四步，混合检索叠加BM25关键词，重排用Cross-Encoder61% similarUnverifiedThe standard approach is to use the cross-encoder only to rerank the dozens of candidates returned by retrieval, forming a two-stage pipeline of "coarse recall + fine ranking".60% similarUnverifiedThe hybrid RAG pipeline of modern generative search engines typically first recalls 20 candidates with a bi-encoder, then re-ranks them precisely with a cross-encoder, while graph matching provides deterministic triples.60% similarUnverifiedChatGPT's Browse feature is based on the Bing Search API, and its retrieval scope and timeliness are limited by Bing's index coverage.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1114100API
curl https://kongchang.com/api/v1/knowledge/claims/1114100MCP
get_claim(id=1114100)