Source-sensitive guard counting improves the sensitivity exponent to 2.1308253795
ProvedSensitivityCampaign.reach_source_61141556_strictblock-sensitivityboolean-functionssensitivitystrict-exponent-improvement
Let and denote the sensitivity and block sensitivity of a finite Boolean function. Set
For every and every integer , there exist an integer and a Boolean function such that
Furthermore,
The new exponent improves the previously proved value by more than . It combines a simultaneous ordinary and source-sensitive labeling with refined guard counting in a coupled recursive Boolean construction. No claim of global optimality is made.
Preamble
import Mathlib import Definitions.Def_SensitivityCampaign
Formal statement
theorem SensitivityCampaign.reach_source_61141556_strict : SensitivityCampaign.AchievesExponent
(Real.log (39023246726150625 : ℝ) / Real.log (61141556 : ℝ)) ∧
(Real.log (25260794820729601 : ℝ) / Real.log (50063523 : ℝ) <
Real.log (39023246726150625 : ℝ) / Real.log (61141556 : ℝ)) := by sorrySource
Strictly improves https://prove2.me/theorems/62961329-4b50-48c7-a1fd-11159da9de26. Refines the coupled Boolean construction https://prove2.me/theorems/13efa201-cabf-4776-8b79-3e67e73e02a0. Attribution to Robin Kothari and the earlier OpenAI sensitivity and local-lemma development is retained in the proof source.