Appendix B.1, p. 33 — r = Ue + (ℰ_v − ℰb) + Y(b − β*) and ‖r‖₂ ≤ ‖Ue‖₂ + ‖ℰ_v − ℰb‖₂ + ‖Y‖₂‖b − β*‖₂
ProvedSynDigits.Ridge.structural_spliterror-boundlatent-factor-modelp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-strengthenedp2o-v1synthetic-control
Consider the latent factor model of §4.1: there are user embeddings and question embeddings (rows ), a noise matrix , and the human responses are . The new question has embedding and noise , with human responses . Decompose with coefficients and residual . For any , and , let and . Then
This splits the structural error of Theorem 4.1 into the part of the new question not representable by the existing ones, the noise, and the distance from to the population Ridge target.
Formalization Note The identity holds for any vector in place of , so no invertibility hypothesis is needed here.
Differences from the source:
- generalization: Σ̃ = (1/n)𝔼[ỸᵀỸ] PSD, γ̃ = (1/n)𝔼[ỸᵀỸ_v], Ridge penalty λ ≥ 0 → any matrix Sig, any vector gam, any real lam (no hypothesis); reason: the step is an algebraic identity plus the triangle inequality and holds for every value, so it specializes to the paper's Σ̃, γ̃, λ
- encoding: v = Vb + e → v = Vᵀ *ᵥ b + e; reason: V ∈ ℝ^{m×d} has the question embeddings as rows and b ∈ ℝ^m, so only Vᵀb type-checks; the proof's Yb = UVᵀb + ℰb (p. 32) confirms it
- reading: Y_v = Uv (§4.1, p. 9, noiseless); ℰ_v appears in Theorem 4.1 without a definition → hYv : Yv = U *ᵥ v + Ev; reason: Theorem 4.1 is stated "in the noisy setting" (p. 11) and its proof writes Y_v − Yb = Uv + ℰ_v − UVb − ℰb (p. 32), so Y_v = Uv + ℰ_v; Ev = 0 recovers the noiseless model of p. 9
- encoding: vectors in ℝⁿ, n×m matrices, ‖·‖₂, spectral norm, σ_min → Fin n → ℝ, Matrix (Fin n) (Fin m) ℝ, vnorm, GaussianMatrix.specNorm, GaussianMatrix.sMin; reason: Lean representation of the same objects; same instances
Preamble
import Mathlib import Definitions.Def_GaussianMatrix_basic import Definitions.Def_SynDigits_Ridge_Model open Matrix GaussianMatrix
Formal statement
namespace SynDigits.Ridge
theorem structural_split {n m d : ℕ} (U : Matrix (Fin n) (Fin d) ℝ)
(V : Matrix (Fin m) (Fin d) ℝ) (Eps Y : Matrix (Fin n) (Fin m) ℝ) (v e : Fin d → ℝ)
(Ev Yv : Fin n → ℝ) (b : Fin m → ℝ) (Sig : Matrix (Fin m) (Fin m) ℝ) (gam : Fin m → ℝ)
(lam : ℝ)
(hY : Y = U * Vᵀ + Eps) (hYv : Yv = U *ᵥ v + Ev) (hv : v = Vᵀ *ᵥ b + e) :
resid Y Yv Sig gam lam = U *ᵥ e + (Ev - Eps *ᵥ b) + Y *ᵥ (b - betaStar Sig gam lam) ∧
vnorm (resid Y Yv Sig gam lam)
≤ vnorm (U *ᵥ e) + vnorm (Ev - Eps *ᵥ b)
+ specNorm Y * vnorm (b - betaStar Sig gam lam) := by sorry
end SynDigits.Ridge
Source
Fan, Huang, Peng, Wang, Wu, SYN-DIGITS: A Synthetic Control Framework for Calibrated Digital Twin Simulation, arXiv:2604.07513v1, p. 33, Appendix B.1, proof of Theorem 4.1, "So, r = …" (with the identity Y_v − Yb = Ue + ℰ_v − ℰb of p. 32)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.