Unverified50% confidenceFactExact time
The weight matrix of the block-diagonal MLP has a block-diagonal structure, which is equivalent to partitioning the input features into several sub-groups and applying linear transformations to each separately, reducing the parameter count from O(n²) to O(n²/b) (b is the number of blocks).
1
Sources
50%
Confidence
Long-term
Relevance
11/9/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe matrix exponential e^{tA} has a series definition for any matrix A, and for the Koopman matrix it can be computed efficiently and numerically via eigendecomposition or the scaling-and-squaring algorithm.69% similarUnverified格是由一组线性无关的基向量通过整数线性组合所张成的离散点集合66% similarUnverifiedThe Partition Principle states: if there exists a surjection from set A to set B, then the cardinality of B is no greater than the cardinality of A.66% similarUnverifiedThe capacitated cluster editing problem takes as input a complete signed edge graph, and the goal is to construct a set of disjoint cliques of size no greater than a given k such that the total number of negative edges inside cliques plus positive edges outside cliques is minimized.65% similarUnverifiedBD partitions the sequence into blocks, uses masked diffusion generation within blocks, and maintains an autoregressive or parallel structure between blocks.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1150261API
curl https://kongchang.com/api/v1/knowledge/claims/1150261MCP
get_claim(id=1150261)