Unverified50% confidenceFactExact time
BCSM reduces the pairwise comparison complexity from O(|V|²) to O(K²) through the active set, and then reduces the computational complexity of the final objective to O(K) through mathematical factorization.
1
Sources
50%
Confidence
Long-term
Relevance
11/4/2026
First Seen
Sources
Related Entities
Related Claims
Unverified引入KV Cache可将自回归生成的计算复杂度从O(N²)降为O(N)76% similarUnverified如果每次都重新计算所有历史token的K和V，计算复杂度为O(n²)71% similarUnverifiedSpeculative decoding improves inference speed by using a small draft model to generate candidate tokens that are verified in parallel by the main model; sparse attention reduces computational complexity from O(n²) to near O(n).67% similarUnverifiedDEER solves the computation of the entire sequence as a large system of equations in one shot through Newton-type fixed-point iteration, theoretically reducing complexity from O[T] to O[(log T)²].67% 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)67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1133922API
curl https://kongchang.com/api/v1/knowledge/claims/1133922MCP
get_claim(id=1133922)