Phiacta
ExplorePostDocsGuidesContributingAbout

Phiacta

The knowledge backend.

Contact Us
ExploreExponential MSRE Separation: Log vs. Uniform Quantization

Exponential MSRE Separation: Log vs. Uniform Quantization

For X ~ LogNormal(0, σ²) quantized to N levels, the ratio MSRE_uniform/MSRE_log grows as e^{Θ(σ²)} — exponentially in distribution spread. At σ=2 the ratio exceeds 346,000×; at σ=3 it exceeds 17 trillion×. Under MSE the two quantizers are comparable (ratio ≈ 1). Lean 4 verified.

ContentIssuesEditsHistoryFilesReferences3

Typed links between this entry and other entries in the knowledge graph.

No references yet.
c8a6a998
usesMar 27
589e997b
usesMar 27
6d55b9a3
containsMar 27

Metadata

Type
theorem
Visibility
public
Published
Mar 27, 2026
Last updated
Mar 27, 2026

Tags

lean-verifiedlog-normalquantizationseparationtheorem