vibe-qcDocumentation

Whitening with a threshold per transfer

AICCM/CCM remains experimental. VibeQCAICCM/WhiteningTransfer.lean adds 17 claims. They extend the whitening statements of Whitening.lean (derivation) from one metric threshold shared by all momentum transfers to a threshold per transfer.

Why

The range-separated fit is the three-dimensional default of the program and the fit the papers report (author decision of 2026-11-14). It decides its metric rule per transfer: every mode is kept when the Cholesky factorisation succeeds and the smallest eigenvalue exceeds the rounding floor, and otherwise the strict cut lambda > tau applies (aux_basis.py::_rsgdf_select_metric_decomposition, _rsgdf_metric_keep). The earlier claims assume one tau for all transfers and certify the classic fit only.

Statements

p is a transfer and m its conjugate partner. “pair” is the unchanged partner hypothesis of cut_folded_tensor.

Claim

Statement

cutq_folded_tensor

The folded-tensor identity with whiteRows (tau q) (e q) (t q), for tau : Q -> R, when partners share their threshold.

cutq_folded_tensor_retained

The same under the weakest condition: partners retain the same modes.

cutq_uniform

With a constant threshold this is cut_folded_tensor.

cutq_full_rank, cutq_full_weight, cutq_full_rows, cutq_full_gram

A threshold below every eigenvalue keeps every mode, and the Gram is the full inverse. This is the Cholesky branch.

cutq_branch, cutq_branch_fold, cutq_branch_rows, cutq_branch_gram

The program’s rule: per transfer either the full-rank rows or the strict cut. The folded-tensor identity holds when partners take the same branch.

cutq_partner_spectrum

A square isometric intertwiner makes partner spectra coincide, so partners need not share an artificial threshold.

cutq_weight_congr, cutq_rows_congr, cutq_whiten_covariant, cutq_matched_scatter, cutq_partner_gram

Per-frame versions of the earlier transport statements.

What is not covered

  • That the program takes the same branch at a transfer and at its conjugate partner. That is the hypothesis share. The branch is a floating-point decision.

  • The short-range and long-range integral evaluation of the range-separated fit, and its cutoff planner.

  • Stability of the retained set near a threshold.

Check

python formal/check_proofs.py with the pinned Lake.