Unverified50% confidenceFactExact time
In a system of dimension d, fully describing the density matrix requires d²−1 real parameters.
1
Sources
50%
Confidence
Long-term
Relevance
10/16/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedWhen the data distribution is supported on a d0-dimensional smooth submanifold, the Wasserstein-p dimension d_p* is typically of the same order as d0, which is much smaller than the ambient dimension D.69% similarUnverifiedDensity functional theory (DFT) predicts material properties by solving the electron density distribution, and its computational complexity grows cubically or even to higher order with the number of electrons; a single calculation on a unit cell containing dozens of atoms may take hours to days.67% similarUnverifiedThe rough estimation formula for the memory (GB) required by a 4-bit quantized model is: parameter count (B) × 0.5 + 2.63% similarUnverified低秩分解将m×n权重矩阵近似为A(m×r)与B(r×n)的乘积，存储参数量从m×n降至r×(m+n)61% similarUnverified2-bit quantization represents each parameter with 2 binary bits, achieving an 8x compression ratio compared to FP16, making it the highest compression tier among currently practical quantization methods, and it is generally only suitable for large-parameter models with redundant structures such as MoE59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1046133API
curl https://kongchang.com/api/v1/knowledge/claims/1046133MCP
get_claim(id=1046133)