Unverified50% confidenceFactExact time
In real-world RAG pipelines, switching between the top 15 MTEB-ranked embedding models produces performance fluctuations of 15 to 25 percentage points.
1
Sources
50%
Confidence
Long-term
Relevance
10/13/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn a real RAG pipeline, merely swapping the embedding model can make recall jump by 22 percentage points.74% similarUnverified相比默认调度策略，块调度可以将系统整体利用率提升15%-25%67% similarUnverifiedFastEmbed-rs所使用的量化模型在MTEB基准测试中相比全精度模型的性能下降通常控制在1-2%以内64% similarUnverified研究表明，即使在 GSM8K 或 MATH 数据集上接近满分的模型，面对稍作变形或需要跨领域组合的问题时性能下降可达30-50个百分点64% similarUnverifiedThe LoRA baseline used for comparison in the paper uses rank 19 with 6.6 million parameters, achieving a correction rate of 83.3%, but exhibited capability loss of 30% to 75%.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1021924API
curl https://kongchang.com/api/v1/knowledge/claims/1021924MCP
get_claim(id=1021924)