Skip to content

Prove exact Clifford Gram endpoints at every odd width - #7788

Open
aliromman wants to merge 1 commit into
CrocSwap:mainfrom
aliromman:codex/consecutive-result-095
Open

aliromman wants to merge 1 commit into
CrocSwap:mainfrom
aliromman:codex/consecutive-result-095

Conversation

@aliromman

@aliromman aliromman commented Nov 19, 2026 •

Copy link
Copy Markdown
Contributor

Enabling algebraic theorem: two explicit degree-eight polynomials give the exact smallest and largest normalized four-qubit Clifford Gram eigenvalues at every odd width f >= 13. The extreme blocks swap with f mod 8. This replaces a fixed sufficient stability constant with the exact width-dependent coefficient-recovery contract. No new kappa or physical decoder is claimed.

Read the all-width proof and two polynomials, complete spectrum and interval certificates, checker, source pins and credits.

Set p=(f-1)/2, sigma=(-1)^floor((f+1)/4) and x=sigma*2^(-p). In the proof, P is the quotient row-one polynomial and Q is row two.

Width modulo 8 Exact minimum Its eigenspace dimension Exact maximum Its eigenspace dimension
1 or 7 Q(x) 313,600 P(x) 256
3 or 5 P(x) 256 Q(x) 313,600

Every other irreducible block lies strictly between these endpoints. The Gram condition number is maximum/minimum; the normalized tensor-map condition number is its square root; the coefficient-recovery norm is the reciprocal square root of the minimum. The full alphabet attains each bound. Principal submatrices give valid bounds for every subalphabet, without a subset sharpness claim.

The complete 81-block spectrum is polynomial in x. There are 318 strict endpoint comparisons on the full intervals [-1/64,0) and (0,1/128]. Dividing each difference by a positive interval coordinate gives 2,544 strictly positive Bernstein coefficients. Two additional certificates prove positive minima. Exact coefficient expansion validates every certificate. This proves the infinite width range; the 35 finite width checks are corroboration. The condition numbers have sharp leading terms 1 + 408*2^(-p) + O(2^(-2p)) for the Gram and 1 + 204*2^(-p) + O(2^(-2p)) for the tensor map.

This extends PR #7772's exact width-thirteen spectrum, head 1398fbdca3536389d8ac3c8b3aaa408f03e8f6c2. It uses the phase and quotient framework from PR #7766, head d7c2dd1868c70f0addce2c0a4d3220149452f14b. The actual phase is negative at widths 13/19 and positive at 15/17. A fixed negative sign would describe the wrong representation at half the odd widths.

Use Python 3.12 and the standard library:

python3.12 -B scripts/research_package.py --package research/clifford-odd-width-spectrum

Both modes print this prefix and elapsed seconds:

PASS_CLIFFORD_ODD_WIDTH_SPECTRUM 81 blocks 318 order proofs 2 positive endpoints

The unchanged threshold source checker runs. Both complete spectral moment identities hold coefficient by coefficient. Multiplicity integrality, dimensions and independent tuple-orbit norms pass at all odd widths 13–81. A literal 24-element one-qubit group checks all 576 products, 384 Pauli quotient fibres and four complete Gram spectra covering every odd residue modulo eight. Six altered-data/domain controls and the separate unsigned-phase control reject.

All 22 retained files are authenticated. Normal/optimized registered replay, syntax/indentation, committed bytes/source history/modes, export and sealed read-only Git-free serial replay pass. Hosted CI did not certify this package. The plan job failed in the shared test_current_ci_section_matches_selected_record: the base guide names frontier6030 while the selected record requires make frontier7000-verify. The log reports 760 tests, one failure and 24 skipped; package replay was skipped. This package changes no shared CI file. Local replay remains the scientific evidence.

Ali Romman (@aliromman) supplied the threshold source and the width-thirteen normalization and stability proof. Nick Petalas (@petalas) supplied PR #7747's Sp(8,2) projection. alephnan (@alephnan) supplied the Pauli-sector and projective Gram framework in PR #6714. The retained notices preserve the ATLAS, Thomas Breuer's CTblLib, MatShep and the earlier framework credit.

The published ordinary character table remains a mathematical input; its archive is not re-extracted here. No full four-qubit Gram matrix, full three-qubit or even-sector suite, selected cold replay, Lean or Docker run is supplied. Native preparation, inverse realization, precision allocation, routing, owned recovery, stock and complete caller costs remain unknown. Exact even-width endpoints are outside scope. No independent human or proof-kernel acceptance is claimed. T(4)=13, all five framework assumptions and the selected frontier remain unchanged.

Base: canonical main e54523517d0f03e58352e647c403a5af3e525a57, priority revision 8. Prepared by Ali Romman with substantial OpenAI Codex assistance in research, implementation, verification and writing. Apache-2.0. Canonical claim.

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.

1 participant