The planted log-likelihood mean lies below the reveal threshold
OpenZeroFreeRRG.Hardcore.overlap_likelihood_mean_thresholdUnder Condition 1 for either positive-parameter model, for every one can choose a reveal probability so that, for all sufficiently large admissible , the planted expectation of the pinned log-likelihood is at most . This is precisely the mean bound with used in the proof of the exponential overlap bound.
Preamble
import Mathlib import Definitions.Def_ZeroFreeRRG_Hardcore_BetheFunctional import Definitions.Def_ZeroFreeRRG_Hardcore_Moments set_option autoImplicit false open Set Filter MeasureTheory AFIsingKS ZeroFreeRRG.Hardcore open scoped Classical BigOperators
Formal statement
theorem ZeroFreeRRG.Hardcore.overlap_likelihood_mean_threshold (m : Model) (hm : m.Valid) (d : ℕ) (hd : 3 ≤ d) (t : ℝ) (ht : m.PosParam t)
(hC1 : Condition1 m d t) (η : ℝ) (hη : 0 < η) :
∃ s : ℝ, 0 < s ∧ s < 1 ∧ ∃ N : ℕ,
∀ n ≥ N, Even (n*d) →
plantedPinnedE m t s d n (fun M σ U => pinnedLL m t M σ U) ≤ (s*η^2/96)*n := by sorrySource
Chen–Wang–Zhang, Zero-Freeness for Random Regular Graphs up to the Replica Symmetry Breaking Threshold, https://github.com/sigact-stoc/stoc2027/releases/download/papers-1/1461.pdf, p. 39, proof of Lemma 5.6.