Novelty audit
=============

The repository includes a source-by-source audit:
`NOVELTY_AUDIT.md <NOVELTY_AUDIT.md>`__.

The generic crossing-number route gives only :math:`\exp(O(N^2))`. However,
Gros--Ramirez Alfonsin's 2025 strong-geometry reconstruction theorem already
implies a factorial-scale upper bound by standard sign-pattern counting. The
full wedge table gives :math:`N^{33N+o(N)}`; using only the signs needed in its
proof gives :math:`N^{6N+o(N)}`. These bounds are implicit consequences rather
than printed corollaries and are credited in the manuscript.

The new direct self-intersection-discriminant argument gives
:math:`N^{3N+o(N)}`, improving the proof-extracted leading upper exponent from
:math:`6` to :math:`3`. The explicit Garside construction gives the lower
exponent :math:`2/3`, while earlier Malyutin--Stupakov and Huh--Oh results
already implied a smaller positive factorial exponent.

The defensible contribution is therefore the sharper direct
topology-to-algebraic-chambers count, together with the new lower construction
and matching order of growth—not a claim that no factorial-scale bound was
previously implicit.