Skip to content

Separate the m=73 log_q stock coefficient 34 from bit length 2642. - #33

Draft
unpaidmoon wants to merge 1 commit into
mainfrom
cursor/logq-stock-shape-gap-387b
Draft

unpaidmoon wants to merge 1 commit into
mainfrom
cursor/logq-stock-shape-gap-387b

Conversation

@unpaidmoon

Copy link
Copy Markdown
Owner

Change and mathematical scope

Finite certificate, scope refined, no new κ. The selected conditional κ stays 0.001672660197076918.

Let C = |Sp(72,2)| * 14826 be the selected m=73 finite stock from research/selected-stock-envelope-m73, with |Sp(72,2)| = 2^1296 * prod_{j=1..36} (2^(2j) - 1) and 14826 = 14336 + 480 + 8 + 2. This package proves

2^2641 <= C < 2^2642

and, for every integer q > 2^80,

log_q C < 34.

So A = 0 and B = 34 satisfy the coarse-depth slot log_q C(w) <= A*w + B at every atom width w >= 1. The integer B = 33 does not: the prime q = 2^80 + 13 has C > q^33 and C <= q^34.

research/selected-stock-shape-a-zero already set A = 0 and substituted the bit length 2642 into

sigma <= (43/26) * (1 - s) * 2642 + rho.

2642 remains a valid log_q upper bound (log_q C < log2(C) < 2642 for every integer q > 2). It is not necessary on q > 2^80. The C-part ratio before the unbound leaf constant rho is exactly 2642/34 = 1321/17.

rho stays unbound, so this is not a numerical sigma. No cutoff margins are recomputed. No supplier, caller, or selected frontier changes. The ordinary BIT coefficients 8194, 6748 and 6741 are a different producer. Historical 57604 and 252 are not used.

Task and overlap

Base bb4bccbf42844d2801487450727c2f63e10d1d0f (upstream main at PR CrocSwap#3725). Task card selected-stock-envelope, priority revision 6, method 2: an exact missing implication for a proposed selected-stage stock envelope. Scope status is refined: the shape note already plugged bit length 2642 into a log_q slot, and this package separates that plug-in from a sufficient log_q coefficient. It does not rederive the bit length or a reserve recurrence.

Session item 28 of the ordered eligible list (eight still-open chronological handoffs, then catalogue methods in file/weight order). Items 1–10 were left for sibling sessions. Item 28 is catalogue method 20.

Nearest predecessors, retained and not replaced:

Open row-ledger notes such as CrocSwap#3667 and CrocSwap#3766 also leave rho unbound. This package does not rerun those ledgers. No open PR found that proves B = 34 on q > 2^80 or the prime 2^80 + 13.

Validation

Environment: Python 3.12, no extra requirements. Author-checked finite certificate. Primality of 2^80 + 13 is Pocklington, not a formal kernel and not an OpenSSL prime call. The factored part of each predecessor is the whole predecessor. Small prime factors are trial-proved through their square roots. Exponent comparisons are integer powers: (2^80 + 1)^34 > 2^2720 > C, and C > (2^80 + 13)^33.

python3.12 -B scripts/research_package.py --package research/logq-stock-shape-gap
python3.12 -I -B research/logq-stock-shape-gap/verify.py
python3.12 -I -O -B research/logq-stock-shape-gap/verify.py
python3.12 -I -B research/logq-stock-shape-gap/independent.py

verify.py exits by SystemExit rather than assert, so -O does not disable the checks. independent.py rebuilds C by multiplying the odd symplectic factors first and does not import verify.py. Expected lines include PASS_LOGQ_STOCK_SHAPE_GAP, PASS_LOGQ_STOCK_SHAPE_INDEPENDENT, and PASS_REGISTERED_RESEARCH_PACKAGE research/logq-stock-shape-gap outcome=completed.

Not run: selected cold replay, Lean, Docker, hosted CI, and any cutoff-margin table.

Credit

The symplectic formula, the family split 14826, and the bit length 2642 are from research/selected-stock-envelope-m73. The log_q slot is from research/selected-coarse-depth-stock. The substitution of 2642 is the composition sentence in research/selected-stock-shape-a-zero. Those packages retain their own predecessor credit. The base-change certificate and the prime 2^80 + 13 are new.

Prepared by unpaidmoon with substantial Cursor Grok assistance. Apache-2.0. No independent human or formal-kernel review is claimed.

Open next questions

From research/logq-stock-shape-gap/TASK.json:

  1. Bind rho for this finite stock, or prove it is zero under a completed-leaf interface. The 1321/17 gap scales only the C-part; a large rho can dominate 34 and 2642 alike. Blocked because the retained complex sources still do not state a per-leaf reserve recurrence.
  2. Check one selected stage at the actual atom width w = ceil(e^s). A uniform log_q coefficient is not yet a stage-width envelope. Blocked because rho is missing, so neither coefficient is a closed sigma.

Workflow feedback (optional)

CrocSwap#1200

Open in Web Open in Cursor 

On every integer base above 2^80 the same finite stock meets the coarse-depth
log_q slot with B=34, while the prime 2^80+13 rejects B=33. Rho stays unbound
and the selected kappa is unchanged.

Co-authored-by: unpaidmoon <unpaidmoon@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants