Theorem The supported later-edge difference is the plane weight
✓ theorem proved

The supported later-edge difference is the plane weight

Fix finite-generator real Coxeter systems \(W,W'\) and an interval order isomorphism \(\varphi:[u,b]_W\simeq[u',b']_{W'}\). Choose a source plane basis \(B\) for \(P\), its induced group \(D\), and a least coset representative \(a_0\): \[ \ell_W(a_0)\le\ell_W(da_0)\qquad(d\in D). \] Use the finite coset graph \(K=\{d\in D:da_0\le_Wb\}\) with its intrinsic order and reflection edges.

Let \(e:x\to_Dy\) be an edge of \(K\) with \(u\le_Wxa_0\). Let \(a\) be its actual positive source root, and let \(\alpha'_e\) be the positive target root of \(\varphi(xa_0)\to_{W'}\varphi(ya_0)\). Let \(F_e,H_e\) be the coset later-edge selections at \(x,y\), using the common threshold \(\alpha'_e\). Write \(\varepsilon_{W'}\) for the ambient after-indicator of the target system.

For a model center \(c\in K\) with \(y\le_Dc\), define the supported selections \[ F_{e,c}=\{f\in F_e:t(f)\le_Dc\},\qquad H_{e,c}=\{f\in H_e:t(f)\le_Dc\}. \] Here \(t(f)\) is the target vertex. Then \[ |F_{e,c}|=|H_{e,c}|+\bigl(k_P(xa_0,a)\bigr)_{\mathbb N}. \] The last term is the natural-number value of the nonnegative geometric plane weight.

Relation to the OpenAI notes

The transported count in the actual model edge quotient.

OpenAI notes: In a supported model, the notes identify the two localized image ideals through their valuations. The later-edge counts differ by the plane weight, which is also the valuation of the correction factor. dih:count, edge:bound.

Lean (primary): The exact supported selections are counted by finite target-vertex sums. The all-ranks transported count and planeWeight_coset_edge identify their difference. Each label and the fixed correction lift are then compared up to units in the actual quotient by cosetLabel e, and structure_scaledImages_of_count gives the image equality. An unsupported model has a zero edge module. PlaneLocalization:1175, TerminalKernels:146.

Role in the proof: The Blueprint checks the selection, root-inequality, pairwise-relative-primality, quotient and count hypotheses separately, then multiplies the unit associations to obtain the exact embedded equality.

Proof

Step 1. Put the current edge in the relative interval with top c

The edge gives \(x\le_Dy\), and the support assumption gives \(y\le_Dc\). Thus \(x\le_Dc\) and \(y\) is a vertex of \([x,c]_D\).

The source bound is \(u\le_Wxa_0\). Since \(c\in K\), the upper bound is \(ca_0\le_Wb\). These are exactly the ambient bounds needed to evaluate \(\varphi(za_0)\) for every \(z\in[x,c]_D\).

Step 2. Express both cardinalities over the same interval

Apply the supported-edge counting theorem with lower interval endpoint \(x\), source \(x\), top \(c\), and threshold \(\alpha'_e\). It gives \[ |F_{e,c}|=\sum_{z\in[x,c]_D} \varepsilon_{W'}(\varphi(xa_0),\alpha'_e,\varphi(za_0)). \] Apply the same theorem with source \(y\), keeping the lower interval endpoint \(x\). Its additional hypothesis \(x\le_Dy\) holds by the current edge, so \[ |H_{e,c}|=\sum_{z\in[x,c]_D} \varepsilon_{W'}(\varphi(ya_0),\alpha'_e,\varphi(za_0)). \] The two applications count their selected supported edges by their target vertices.

Step 3. Apply the transported dihedral count

The all-ranks transported-count theorem applies to the setup verified in Step 1 and to the current edge. Its threshold is precisely the positive root \(\alpha'_e\) used by both selections.

It identifies the first sum in Step 2 with the second sum plus \[ \frac{\ell_D(y)-\ell_D(x)-1}{2}. \] The theorem includes the case \(c=y\) and the relative rank-one case.

Step 4. Identify the correction with the geometric weight

The relative edge-degree theorem applies to this same least coset representative and actual source root. It gives \[ \bigl(k_P(xa_0,a)\bigr)_{\mathbb N} =\frac{\ell_D(y)-\ell_D(x)-1}{2}. \] Substitute this into Step 3 and then use the cardinality identities from Step 2. The result is the stated natural-number equality.

References

The local source declarations below give the construction and proof.

Lean formalization

OAI.KLInvariance.TitsSpace.PlaneBasis.supported_count_difference: PlaneLocalization.lean:1175.

Used by