--- served-v2.md 2026-09-25 19:09:33.499389000 +0200 +++ revised-attack-0830-varE-identification.md 2026-09-25 19:13:48.015302800 +0200 @@ -5,7 +5,7 @@ status: PARTIAL todo: 9 question: Can either open step behind lim Var/E = 0.45546 (the identification delta*(X - X_dec) -> 0, or the theta = 2 mean-coefficient replacement) be proven, and if not, which single inequality does not close? -verdict: Neither closes, and the two named steps are one statement (Conjecture 1 of variance-note sec.10; the model half is PROVEN in varE-limit-theorem.md). The remainder is re-split exactly at five levels: the corpus's X2 column doubles the positive half of one shift pattern (X2c = 2.66 against the group sum 0.09 at x = 19, PROVEN by rebuild), so its Xmix is overstated 2.3x; the true mixed remainder is MEASURED at -0.49 ln y and 82 % of it sits BELOW 2L, in the two-branch type, in the unbalanced range min(d,e) <= L^(2/5) (99.7 % of that cell at x = 19), outside the range of every bilinear Kloosterman-fraction bound; the divisor-distribution estimate the record names (n > 2L only) is therefore not the whole open step. Henriot's Corollary 2 (read at the page, erratum read) applies as stated and cuts the moduli to n <= L ln^(2+o(1)) y, PROVEN given that theorem; the one inequality left is the uniform o(1) equidistribution of y-friable squarefree integers in progressions to moduli up to y^(4/5), which has since been searched at the page in recon-0830-smooth-aps.md (Q-recon-0830-smooth-aps, same TODO item 9): eleven sources, NONE APPLIES AS STATED; per that note's 2026-08-30 rider from redteam-0830-imports.md the nearest statement is Harper, JLMS 112 (2025) e70293 = arXiv:2412.19644 (the BDH asymptotic for an arbitrary complex sequence on the same dyadic all-classes object), and the step still does not close, on the range sqrt(2x) < Q <= x and because its Theorem 2 route bars sieved sets by name. The limit 0.45546 stays HEURISTIC. Not TPC-strength. +verdict: Neither closes, and the two named steps are one statement (Conjecture 1 of variance-note sec.10; the model half is PROVEN in varE-limit-theorem.md). The remainder is re-split exactly at five levels: the corpus's X2 column doubles the positive half of one shift pattern (X2c = 2.66 against the group sum 0.09 at x = 19, PROVEN by rebuild), so its Xmix is overstated 2.3x; the true mixed remainder is MEASURED at -0.49 ln y and 82 % of it sits BELOW 2L, in the two-branch type, in the unbalanced range min(d,e) <= L^(2/5) (99.7 % of that cell at x = 19), outside the reach of the bilinear Kloosterman-fraction bounds in print below every fixed power L^delta and in the three-branch / sharp-cut complements (DFI97 Thm 1, whose Fejer-weighted rectangular block (A) is PROVEN in return #916, reaches from L^delta up to L^(2/5)); the divisor-distribution estimate the record names (n > 2L only) is therefore not the whole open step. Henriot's Corollary 2 (read at the page, erratum read) applies as the erratum states it, under eps < alpha/600 and with F_0, F_1 read at prime powers (redteam-0830-imports.md row a5), and cuts the moduli to n <= L ln^(2+o(1)) y, PROVEN given that theorem; the one inequality left is the uniform o(1) equidistribution of y-friable squarefree integers in progressions to moduli up to y^(4/5), which has since been searched at the page in recon-0830-smooth-aps.md (Q-recon-0830-smooth-aps, same TODO item 9): eleven sources, NONE APPLIES AS STATED; per that note's 2026-08-30 rider from redteam-0830-imports.md the nearest statement is Harper, JLMS 112 (2025) e70293 = arXiv:2412.19644 (the BDH asymptotic for an arbitrary complex sequence on the same dyadic all-classes object), and the step still does not close, on Harper's range sqrt(2N) < Q <= N (N the length of his sum) and because its Theorem 2 route bars sieved sets by name. The limit 0.45546 stays HEURISTIC. Not TPC-strength. --> *Staging note, 2026-08-30. TODO item 9, the "prove either step" move. Producer: @@ -86,13 +86,19 @@ branch, `n <= 2L`) split by balance: the pairs with `min(d, e) <= L^(2/5)` carry `95.7, 93.2, 94.4, 98.7, 99.7 %` of the cell's remainder at `x = 7..19`; the balanced pairs' remainder reads `-0.0087` at `x = 19`. The - obstruction is unbalanced, and unbalanced is exactly where no Kloosterman- - fraction bound in print reaches (section 4). + obstruction is unbalanced, and the Kloosterman-fraction bounds in print reach + only part of the unbalanced range: DFI97 Theorem 1 (statement via Shen + 2607.06575 Lemma 5) saves `D^{-1/2} + sqrt(D/E)` for a fixed-power short + factor `L^delta <= D <= L^{2/5}`, and return #916 (accepted at proven) proves + its Fejer-weighted rectangular block (A), so what no bound in print reaches is + a short factor below every fixed power `L^delta`, together with the + three-branch and sharp-cut complements of the cell (section 4). - **PROVEN given a theorem read at the page (section 5).** Henriot's Corollary 2 (`arXiv:1102.1643v1`, page 7; erratum MPCPS 157 (2014) 375-377 read: "the upper bounds in Theorem 5 and Corollaries 1-2 are not sharp as claimed - (although they are still valid)") applies to `F_0(h) F_1(h-2) F_1(h+2)` with - its hypotheses met as stated, and gives + (although they are still valid)") applies to `F_0(h) F_1(h-2) F_1(h+2)` as the + ERRATUM states it, under `eps < alpha/600` and with `F_0, F_1` read at prime + powers (`redteam-0830-imports.md` row a5; section 5), and gives `sum_{|h| L ln^2 y * phi(y)`, any `phi -> infinity`, contribute `o(ln^2 y)` to both the count and the main term. The open estimate's range shrinks from @@ -151,7 +157,10 @@ band by band, and PART C reads the `n <= 2L` band of the mixed remainder at `-3.22` against `-0.72` above `2L` at `x = 19`. Both bands are one logarithm short under the kernel bound (`sum_{n<=2L} W_mix(n) n/L` and -`sum_{n>2L} W_mix(n)` are both `O(ln^3 y)` by the `4^omega` count), so the +`sum_{n>2L} W_mix(n)` are both `O(ln^3 y)` by the `4^omega` weighted coefficient +mass (the sum of `2^{#zero labels}` over the assignments of labels to the primes), +not by a class count of `4^omega`: the classes number `3^omega` (return #913 +section 4)), so the `n <= 2L` band is not a smaller problem; it is the same problem at a different scale, `r_n/n` with `r_n = x# mod n` in place of `L/n`. @@ -243,10 +252,15 @@ balanced cell's remainder at `-0.0087` at `x = 19`, so a power saving there is at least consistent with the data. 5. *The unbalanced range, which is the obstruction.* PART G: `99.7 %` of the - dominant cell sits at `min(d, e) <= L^{2/5}`, and no bound in print for - `e(a inv(m)/n)` with arbitrary coefficients saves anything when one variable - is below the `2/5`-power of the other in this configuration. Section 4 - states what is needed there. + dominant cell sits at `min(d, e) <= L^{2/5}`. For `e(a inv(m)/n)` with + arbitrary coefficients a bound in print does save in part of that range: + DFI97 Theorem 1 (statement via Shen 2607.06575 Lemma 5, and review #184 on + return #916) gives `D^{-1/2} + sqrt(D/E)` for a fixed-power short factor + `L^delta <= D <= L^{2/5}`, and return #916 (accepted at proven) proves its + Fejer-weighted rectangular block (A), which is the shape of the cell here. + What no bound in print reaches is a short factor below every fixed power + `L^delta`, and the three-branch and sharp-cut complements. Section 4 states + what is needed there. ## 4. The single inequality that does not close @@ -319,10 +333,11 @@ |---|---|---|---| | Shiu 1980, Crelle 313 | quoted at second hand from Nair-Tenenbaum p.119 (read at the text layer of the Acta Math. PDF, sha256 `3a2caa93...`): `sum_{x= y >= x^alpha`, `q <= y^{1-beta}` (the last range read through OCR damage; not load-bearing) | one multiplicative function of one variable; ours is a product of three functions on three shifted arguments | HYPOTHESIS UNMET as stated; superseded by the next two rows | | Nair-Tenenbaum 1998, Acta Math. 180, Theorem 1 (p.125, read) | `sum_{x= c_0 \|\|Q\|\|^delta`: `<< Delta_{D*} y prod_{g= c_0 \|\|Q\|\|^delta`: `<< Delta_{D*} y prod_{g= L^delta` is needed | our coefficients are arbitrary and coprime-supported and the phase is the right one (PART A); its Fejer-weighted rectangular block (A) is the shape of the dominant cell here (PART G) | **HYPOTHESIS MET for block (A)**, `\|a\| << MN` and `L^delta <= D <= L^{2/5}` (PROVEN in return #916); the three-branch type and the sharp-cut complements are NOT of shape (A); below every fixed power `L^delta` it does not reach the obstruction | +| Bettin-Chandee Theorem 1 (read, p.2) | bilinear / trilinear forms with `e(a inv(m)/n)`, arbitrary coefficients, saving a power when the two variables are balanced | our coefficients are arbitrary and coprime-supported, the phase is the right one (PART A); the kernel couples the third variable to the product | HYPOTHESIS MET on the balanced blocks; owed the separation; does not reach the obstruction | **The deduction from Henriot's Corollary 2 (PROVEN given the theorem as read).** Apply it on each dyadic block of `0 < h < L` with `y = x` (and by symmetry on