Unverified50% confidenceFactExact time
The naive discrete Fourier transform (DFT) algorithm requires O(N²) complex multiplications.
1
Sources
50%
Confidence
Long-term
Relevance
10/13/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedReverse-mode automatic differentiation can compute the gradients of a scalar loss with respect to all parameters in O(n) time, whereas forward mode requires O(n) forward passes71% similarUnverifiedL2算法相对于基向量系数位长β达到O(n⁴·β·(n+β))的位复杂度，在β=O(n)时相当于O(n⁶)，而经典LLL约为O(n⁹)70% similarUnverifiedLow-rank parameterization decomposes the full covariance matrix into the form Σ = WW^T + D, reducing the parameter count from O(d²) to O(dr).65% similarUnverified引入KV Cache可将自回归生成的计算复杂度从O(N²)降为O(N)65% similarUnverifiedOU过程具有均值回归性质，其数学形式为 dXt = -αXt dt + σ dWt，通过适当设置参数任意初始分布会在有限时间内指数级收敛到标准高斯分布64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1013570API
curl https://kongchang.com/api/v1/knowledge/claims/1013570MCP
get_claim(id=1013570)