Unverified50% confidenceFactExact time
In a real RAG pipeline, merely swapping the embedding model can make recall jump by 22 percentage points.
1
Sources
50%
Confidence
Long-term
Relevance
10/13/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn real-world RAG pipelines, switching between the top 15 MTEB-ranked embedding models produces performance fluctuations of 15 to 25 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%61% similarUnverifiedAt small scale, the FLOPs share of the MLP layer is approximately 44%, rising to approximately 80% when scaled to 175B parameters58% similarUnverified权重绑定在中小规模模型上(如GPT-2 124M)通常能提升性能并减少约30%的嵌入相关参数，但在超大规模模型中收益递减57% similarVerifiedLoRA decomposes the fine-tuning weight change ΔW into the product of two low-rank matrices, with a parameter count of only 0.1%–1% of the original model, achieving a parameter compression ratio as high as 256× when r=8 and d=4096.56% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1021921API
curl https://kongchang.com/api/v1/knowledge/claims/1021921MCP
get_claim(id=1021921)