theorem
proved
poisson_enhancement
show as:
view math explainer →
open explainer
Read the cached plain-language explainer.
open lean source
IndisputableMonolith.ILG.PoissonKernel on GitHub at line 36.
browse module
All declarations in this module, on Recognition.
explainer page
depends on
formal source
33
34/-- Stability/Scaling Bound: The ILG potential Φ is strictly enhanced relative to
35 the GR potential Φ_GR by exactly the kernel factor w(k, a). -/
36theorem poisson_enhancement (P : KernelParams) (k a δρ : ℝ) (hk : k ≠ 0) :
37 let Φ_ILG := poisson_operator P k a δρ
38 let Φ_GR := -(4 * Real.pi * δρ) / k^2
39 |Φ_ILG| = kernel P k a * |Φ_GR| := by
40 unfold poisson_operator
41 simp only [if_neg hk]
42 have h_kernel_pos : 0 < kernel P k a := kernel_pos P k a
43 -- Rewrite -(4πw·δρ)/k² as w·(-(4π·δρ)/k²) under absolute value.
44 have h_eq : -(4 * Real.pi * kernel P k a * δρ) / k^2
45 = kernel P k a * (-(4 * Real.pi * δρ) / k^2) := by ring
46 rw [h_eq, abs_mul, abs_of_pos h_kernel_pos]
47
48/-- Coercivity Bound: The modified potential is non-vanishing for any non-vanishing source. -/
49theorem poisson_coercive (P : KernelParams) (k a δρ : ℝ) (hk : k ≠ 0) (hδρ : δρ ≠ 0) :
50 poisson_operator P k a δρ ≠ 0 := by
51 unfold poisson_operator
52 simp only [if_neg hk]
53 have hk2 : (k^2 : ℝ) ≠ 0 := pow_ne_zero 2 hk
54 have h4pi_ne : (4 * Real.pi : ℝ) ≠ 0 :=
55 mul_ne_zero (by norm_num) Real.pi_ne_zero
56 have hkern_ne : kernel P k a ≠ 0 := (kernel_pos P k a).ne'
57 have hnum_ne : (4 * Real.pi * kernel P k a * δρ : ℝ) ≠ 0 :=
58 mul_ne_zero (mul_ne_zero h4pi_ne hkern_ne) hδρ
59 have hneg_ne : -(4 * Real.pi * kernel P k a * δρ : ℝ) ≠ 0 := neg_ne_zero.mpr hnum_ne
60 exact div_ne_zero hneg_ne hk2
61
62/-! ## Causality-bound Poisson operators (Beltracchi 2026 resolution)
63
64The original `poisson_operator` above is preserved unchanged. The two
65operators below split the Poisson equation into a background piece
66(unmodified standard GR) and a perturbation piece (ILG-modified, with the