C3.8U · FORMULA WORKBENCH

PN.2 formula workbench

Lower bounds for selected separations. Each formula links to its assumptions and a verified proof page.

Main R/D boundℛ_Rℛ_D≈ 0.375

Parameters

Unit labels

Convert inputs to consistent units first. Labels do not perform unit conversion. [c] = [S]/[R].

Calculation updated. Results are conditional mathematical bounds.

Lower bounds

BoundValue ≈Unit
ℛ_S11
ℛ_R0.51
ℛ_D0.751
ℛ_Sℛ_D0.751
ℛ_Rℛ_D0.3751
ℛ_Sℛ_D (c)0.3751

The numerical condition Δ_S ≥ c‖K‖ holds. This does not prove separation throughout Ω.

Factor (1 − τ)²
Factor (1 − τ)²1000.51τ

Export includes inputs, exact fractions, units, formula references and evidence statuses.

Entering numbers does not validate a physical model. e_X is the maximum error of a fixed estimate; ℛ_X is the maximum expected metric loss. At τ = 0 the coefficients agree, but the definitions remain different.

Formulas and assumptions

K-DEF · Definitions

Associator

K(x,y,z)=(xy)zx(yz)=rLrR K(x,y,z)=(x\odot y)\odot z-x\odot(y\odot z)=r_L-r_R

Objects and domain are specified in a normed algebra; this is a definition, not an experimental result.

SOURCE_ANCHORED · Section 2; 3.1 · PDF page 2

TeX

GAPS · Definitions

Branch separations

ΔS=|S(rL)S(rR)|,ΔD=|𝔇(bL)𝔇(bR)|,ΔR=K \Delta_S=|S(r_L)-S(r_R)|,\quad\Delta_D=|\mathfrak D(b_L)-\mathfrak D(b_R)|,\quad\Delta_R=\|K\|

Objects and domain are specified in a normed algebra; this is a definition, not an experimental result.

SOURCE_ANCHORED · Section 3.4 · PDF page 3

TeX

DET-ERROR · Definitions

Deterministic error

eX=maxb{L,R}dX(x̂,xb) e_X=\max_{b\in\{L,R\}}d_X(\widehat x,x_b)

Theorem package: A1–A4, one estimator for hidden branches. A positive bound requires nonzero separations.

SOURCE_ANCHORED · Section 3.3 · PDF page 3

TeX

C-SEP · Definitions

c-separation condition

ΔS(u)cK(u),c>0,uΩ \Delta_S(u)\ge c\|K(u)\|,\qquad c>0,\quad u\in\Omega

Additionally c > 0 and Δ_S(u) ≥ c‖K(u)‖. A numerical row does not prove separation over all of Ω.

SOURCE_ANCHORED · Section 3.5 · PDF page 3

TeX

PN2-S · Deterministic

PN.2-S · scalar form

eSeDΔSΔD4 e_Se_D\ge\frac{\Delta_S\Delta_D}{4}

Theorem package: A1–A4, one estimator for hidden branches. A positive bound requires nonzero separations.

SOURCE_ANCHORED · Section 7; 8 · PDF page 4

TeX

PN2-R · Deterministic

PN.2-R · full state

eReDKΔD4 e_Re_D\ge\frac{\|K\|\Delta_D}{4}

Theorem package: A1–A4, one estimator for hidden branches. A positive bound requires nonzero separations.

SOURCE_ANCHORED · Section 7; 8 · PDF page 4

TeX

PN2-K · Deterministic

PN.2-K · c-separation

eSeDcKΔD4 e_Se_D\ge\frac{c\|K\|\Delta_D}{4}

Additionally c > 0 and Δ_S(u) ≥ c‖K(u)‖. A numerical row does not prove separation over all of Ω.

SOURCE_ANCHORED · Section 7; 8 · PDF page 5

TeX

TV · Statistical

Data distinguishability

τ=TV(PL,PR)=12|pq|dμ,μ=PL+PR \tau=\operatorname{TV}(P_L,P_R)=\frac12\int|p-q|\,d\mu,\quad\mu=P_L+P_R

Two probability laws on one measurable space; τ uses the entire accessible observation.

SOURCE_ANCHORED · Section 3 · PDF page 2

TeX

OVERLAP · Statistical

Overlap of laws

ν(𝒴)=min(p,q)dμ=1τ \nu(\mathcal Y)=\int\min(p,q)\,d\mu=1-\tau

Two probability laws on one measurable space; τ uses the entire accessible observation.

SOURCE_ANCHORED · Section 3 · PDF page 3

TeX

STAT-RISK · Statistical

Worst expected risk

X=maxb{L,R}𝔼bdX(X̂,xb) \mathcal R_X=\max_{b\in\{L,R\}}\mathbb E_b\,d_X(\widehat X,x_b)

C3.8T audit: A1–A6, common complete-data laws and one decision kernel; finite risks.

SOURCE_ANCHORED · Section 2 · PDF page 2

TeX

STAT-S · Statistical

Risk S

SΔS2(1τ) \mathcal R_S\ge\frac{\Delta_S}{2}(1-\tau)

C3.8T audit: A1–A6, common complete-data laws and one decision kernel; finite risks.

SOURCE_ANCHORED · Section 4; 5 · PDF page 3

TeX

STAT-D · Statistical

Risk D

DΔD2(1τ) \mathcal R_D\ge\frac{\Delta_D}{2}(1-\tau)

C3.8T audit: A1–A6, common complete-data laws and one decision kernel; finite risks.

SOURCE_ANCHORED · Section 4; 5 · PDF page 3

TeX

STAT-R · Statistical

Risk R

RK2(1τ) \mathcal R_R\ge\frac{\|K\|}{2}(1-\tau)

C3.8T audit: A1–A6, common complete-data laws and one decision kernel; finite risks.

SOURCE_ANCHORED · Section 4; 5 · PDF page 3

TeX

STAT-SD · Statistical

Product of S/D risks

SDΔSΔD4(1τ)2 \mathcal R_S\mathcal R_D\ge\frac{\Delta_S\Delta_D}{4}(1-\tau)^2

C3.8T audit: A1–A6, common complete-data laws and one decision kernel; finite risks.

SOURCE_ANCHORED · Section 5 · PDF page 4

TeX

STAT-RD · Statistical

Product of R/D risks

RDKΔD4(1τ)2 \mathcal R_R\mathcal R_D\ge\frac{\|K\|\Delta_D}{4}(1-\tau)^2

C3.8T audit: A1–A6, common complete-data laws and one decision kernel; finite risks.

SOURCE_ANCHORED · Section 5 · PDF page 4

TeX

STAT-cSD · Statistical

Risks with c-separation

SDcKΔD4(1τ)2 \mathcal R_S\mathcal R_D\ge\frac{c\|K\|\Delta_D}{4}(1-\tau)^2

Additionally c > 0 and Δ_S(u) ≥ c‖K(u)‖. A numerical row does not prove separation over all of Ω.

SOURCE_ANCHORED · Section 5 · PDF page 4

TeX

Assumptions and scope

Deterministic model: normed algebra, two defined bracket outcomes, hidden branch index, one estimate and Δ_D > 0. The c-form additionally requires proven c-separation.

Statistical model: the entire accessible observation Y, laws P_L/P_R, measurable metric loss, one branch-independent decision kernel and finite risks; K = r_L − r_R. The c-form requires separation.

The canonical statistical patch has not been applied. Physical validation has not been performed. truth_layer_promotion=0 · EXT_RUN=NOT_RUN · NO_LOSS_FREEZE=BLOCKED.