Neutralizing Shadow Terms in BPS/Microstate Counts: A Projection Framework and L-Value Transduction (v1.1)

# Neutralizing Shadow Terms in BPS/Microstate Counts: A Projection Framework and L-Value Transduction (v1.1) **Author:** Lee Byoungwoo **Version:** v1.1 ## OverviewThis preprint isolates and removes a persistent obstruction in BPS-state / microstate-count asymptotics driven by a continuum-mode “shadow” contribution (heuristically of the form \(c\,X^{1/2}\)). The core idea is to **neutralize** this mode at the level of test weights by an explicit orthogonalization in Mellin space, and then to connect the remaining (discrete/automorphic) contribution to \(L\)-values through an automorphic projection pipeline. The resulting framework is organized as a reproducible chain: 1. **Windowing:** choose a standard Mellin-aligned window \(\psi_T\) (dyadic or log–Gaussian).2. **Neutralization:** remove the continuum direction \(v(t)=t^{1/2}\) to form \(f_T\).3. **Projection:** use an automorphic projection \(P_{\mathrm{Aut}}\) (Kohnen plus-space / \(\xi\)-operator / Shimura–Niwa–Shintani bridge, depending on the target).4. **Mellin–\(L\) identity:** reduce the discrete piece to \(L\)-data.5. **Transduction:** import an analytic input (subconvexity or mean/hybrid bounds) and output an improved remainder exponent for the microstate observable. A small worked example (index \(m=1\), level \(N=11\)) and a negative control are included to illustrate the measurement logic. --- ## Closed results (what is proved here) ### 1) Neutralization theorem (structural + quantitative tail)Let \(\psi_T\) be a standard dyadic or log–Gaussian window and define\[f_T \;:=\; \psi_T \;-\; \frac{\langle \psi_T,v\rangle_{dt/t}}{\langle v,v\rangle_{dt/t}}\,v,\qquad v(t)=t^{1/2},\qquad \langle f,g\rangle_{dt/t}=\int_0^\infty f(t)\overline{g(t)}\,\frac{dt}{t}.\]Then the observable admits a discrete/continuous decomposition where the continuum main term cancels **exactly** by construction, and the remainder splits into two explicit tails: \[D_X[f_T]\;=\;D_X^{\mathrm{disc}}[f_T]\;+\;O_{\mathcal{P},\varepsilon}\!\big(T^{-3/2+\varepsilon}\big)\;+\;O_{\mathcal{P},\varepsilon}\!\big(X^{-1-2\eta}\big).\] Here \(\mathcal{P}\) is a fixed “parameter box” (window family, bandwidth, truncation interval, and the \(X\leftrightarrow T\) mapping). **Important:** the *structural cancellation* is assumption-free; the *quantitative tail rates* use standard analytic hypotheses (listed as SD1–SD3 in the paper). ### 2) Automorphic projection stabilityUnder a short normalization audit (Archimedean factor matching, Hecke/Atkin–Lehner compatibility, and Mellin-measure convention \(dt/t\)), the projection \(P_{\mathrm{Aut}}\) is bounded on the ambient model space and the Mellin–\(L\) identity is stable under neutralization \(\psi_T\mapsto f_T\). ### 3) Transduction theorems (input exponent \(\to\) output exponent)**Pointwise input:** if\[L\!\big(\tfrac12+it,\pi\big)\ll_\varepsilon t^{\,1/4-\delta_{\mathrm{in}}+\varepsilon},\]then the framework outputs (on dyadic \(X\)-ranges)\[R_\pi(X;T)\ll_\varepsilon X^{\,1/4-\delta_{\mathrm{in}}+\varepsilon}.\] **Mean/hybrid input:** if a mean/hybrid bound gives a gain \(\delta'_{\mathrm{in}}>0\) near \(s=\tfrac12\), then for standard windows there exists an efficiency constant \(c\ge 2/3\) such that the output exponent satisfies\[\beta_{\mathrm{out}}=\tfrac14-c\,\delta'_{\mathrm{in}},\qquadR_\pi(X;T)\ll_\varepsilon X^{\,\beta_{\mathrm{out}}+\varepsilon}.\] --- ## What is new in v1.1- A single “projection + neutralization + transduction” pipeline written in a theorem/protocol format suitable for scrutiny.- Explicit separation of the remainder into \(T\)-tail and \(X\)-tail with parameter-box bookkeeping.- An efficiency statement for standard windows (dyadic/log–Gaussian) yielding \(c\ge 2/3\).- A worked example template (index \(m=1\), level \(N=11\)) and a negative control to stress-test slope claims.- A one-page reproducibility checklist (A1–A8) and a units/Jacobian audit (to prevent unit-dependent “fake gains”). --- ## Scope & non-toy statusThis paper is an **engine**: it does not claim to derive quantum gravity or microstate formulas from first principles. It provides a mechanism to1) remove a universal continuum obstruction at the level of test weights, and 2) convert analytic \(L\)-function input into a sharper remainder exponent for BPS/microstate observables. **What is *not* claimed:**- No proof of the external analytic inputs (e.g., subconvexity) is given here.- No claim is made that the framework alone fixes absolute SI-valued constants; outputs are exponent-level and normalization-audited. **What is falsifiable/checkable now:**- The pre/post slope drop in \(\log D_X\) vs. \(\log X\) under \(\psi_T\mapsto f_T\), plus a negative control, using the same \((T,\sigma,[t_{\min},t_{\max}])\).- Window robustness (dyadic vs. log–Gaussian) within the declared parameter box \(\mathcal{P}\).- Unit-independence of the reported exponents under allowed variable changes (audited in the Units & Jacobian appendix).- Numerical verification of the Mellin–\(L\) identity on a sample set of \(s\) with \(\Re s>1\) (sanity check of the automorphic bridge). --- ## Reproducibility (quick start)- Compile the manuscript: - `pdflatex Neutralizing_BPS_v1.1.tex` - `pdflatex Neutralizing_BPS_v1.1.tex` (run twice for references)- Follow the “Numerical protocol (repro-ready)” section in the paper.- Use the Appendix “Repro checklist (A1–A8)” as a minimal acceptance gate before claiming a slope improvement. Recommended supplementary artifacts (if you attach them to the Zenodo record):- a small run log (CSV) with the chosen \(\mathcal{P}\), windows, fit ranges, and estimated slopes,- a manifest with SHA256 checksums,- environment note (TeX version + OS + any scripts). --- ## Program closure and targets (next steps)- Strengthen or remove SD1–SD3 by proving them for broader kernel classes and/or larger automorphic families.- Expand beyond holomorphic/\(\theta\) targets to Maass targets with a uniform projection audit.- Provide a compact reference implementation (notebook or script) that reproduces the worked example end-to-end, including the negative control.- Map the efficiency constant \(c\) as a function of window design (dyadic vs. log–Gaussian; bandwidth \(\sigma\)) to support “best-practice” defaults. --- ## Keywords / MSC**Keywords:** mock modular forms, harmonic Maass forms, Kohnen plus space, theta lift, Mellin transform, automorphic projection, subconvexity, BPS state counting, black hole microstates, remainder terms, windowing/regularization. **MSC (suggested):** 11F37, 11F50, 11F72, 11F03, 83C57. --- ## Minimal background pointers (non-exhaustive)- Kohnen (1980), *Modular forms of half-integral weight on \(\Gamma_0(4)\)*.- Bruinier & Funke (2004), *On two geometric theta lifts*.- Iwaniec & Sarnak (1995), *Perspectives on the analytic theory of \(L\)-functions*.