Unverified50% confidenceFactExact time
LBFGS is a quasi-Newton method that approximates the inverse of the Hessian matrix by storing gradient difference information from the most recent m steps, requiring only O(md) memory.
1
Sources
50%
Confidence
Long-term
Relevance
9/3/2026
First Seen
Sources
Related Entities
Related Claims
Unverified典型的导纳模型可表达为 M·ẍ + B·ẋ + K·x = F_ext，其中M、B、K分别是期望的虚拟惯量、阻尼和刚度矩阵60% similarUnverifiedGPTQ is a post-training quantization scheme based on layer-wise Hessian matrices, AWQ is activation-aware weight quantization, and modern quantization schemes can control perplexity loss to within 1–2%.60% similarUnverified传统线性时不变SSM（如S4模型）系统矩阵A、B、C在序列处理中保持不变，递推关系可展开为卷积形式，利用FFT在O(N log N)时间内并行计算60% similarVerifiedCommon model quantization methods include formats such as GPTQ, AWQ, and GGUF.57% similarUnverified选择性量化思路与LLM领域的GPTQ、AWQ等量化方法一脉相承57% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/850173API
curl https://kongchang.com/api/v1/knowledge/claims/850173MCP
get_claim(id=850173)