Unverified50% confidenceTradeoffExact time
The K value for Top-K precomputation typically falls between 5 and 20, aligned with the number of suggestions displayed in the user interface, and the index size grows linearly with K.
1
Sources
50%
Confidence
Long-term
Relevance
9/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedK折交叉验证的K通常取5或10，这在偏差-方差权衡上被实证为较优选择65% similarUnverifiedThe developer ultimately selected the variant with a KL divergence of 0.028 as the optimal solution.64% similarUnverifiedKimi expanded from K2's 1 trillion parameters to K3's 2.8 trillion parameters.64% similarUnverified在MoE模型中每个token只激活少量Expert（通常2-4个，Top-K路由），实际计算量可能仅相当于6-7B的Dense模型63% similarUnverifiedTop-K SAE retains only the K features with the highest activation values during encoding and forces all others to zero, thereby compelling the model to learn sparse representations.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/911709API
curl https://kongchang.com/api/v1/knowledge/claims/911709MCP
get_claim(id=911709)