KL endpoint formulas under character identification
Fix a finite-generator Coxeter system and its real polynomial coefficient algebra \(A\) in \(N\) degree-one variables. Let \(B\) be any normalized boundary sheaf on \([u,b]\), and let \(x\in[u,b]\). Suppose its actual character satisfies the vertex equality \[ C_x^{[u,b]}(q)=P_{x,b}(q). \] For \(0<t<1\), define the normalized kernel state and scaled KL polynomial by \[ D_x(t,F)=(1-t^2)^N t^{\ell(x)-\ell(b)}H_{L_x(F)}(t^2), \qquad p_{x,b}(t)=t^{\ell(x)-\ell(b)}P_{x,b}(t^2). \] Here \(L_x(F)\) imposes the selected outgoing vanishing conditions and \(H_M(q)=\sum_{n\ge0}\dim_{\mathbb R}M_n\,q^n\).
Then \[ D_x(t,\varnothing)=p_{x,b}(t). \] For the full-kernel assertion, take \(u=1\) and additionally supply section duality with homogeneous pairing shift \(\ell(b)\). Under the same character equality at \(x\), \[ D_x(t,U_x)=p_{x,b}(t^{-1}), \] where \(U_x\) is the full outgoing edge set.
Relation to the OpenAI notes
Different intermediate statement.
OpenAI notes: The grading-normalization proposition identifies the stalk graded rank with the canonical KL polynomial and uses the Elias–Williamson character theorem through the cited moment-graph bridges. bg:normalization.
Lean (primary): The development defines the equality of actual stalk characters and canonical KL polynomials as StalkCharacterObligation. The unconditional theorem used in the main route is R-reciprocity for the actual character. The local strict half-degree implication retains an explicit nonzero-diagonal hypothesis. SheafMorphisms:219, DyerPaths:62, PositiveTransfer:186, CharacterReciprocity:1807, TerminalKernels:1273, TerminalKernels:1284.
Role in the proof: The comparison uses the actual character C and its proved R-reciprocity. Canonical P is handled by the later polynomial-rigidity induction. The local degree result applies to a diagonal coefficient satisfying its stated nonvanishing hypothesis.
Proof
Step 1. Compute the stalk Hilbert value in the supplied model
The given \(B\) is graded-isomorphic to the chosen boundary sheaf used to define \(C_x^{[u,b]}\). Its stalk therefore has that same character.
With no outgoing conditions, \(L_x(\varnothing)=B^x\). The homogeneous free-basis formula gives \[ H_{L_x(\varnothing)}(t^2) =C_x^{[u,b]}(t^2)(1-t^2)^{-N}. \] The range \(0<t<1\) ensures \(0<t^2<1\), as required for this convergent Hilbert value.
Step 2. Substitute the character equality at the empty endpoint
Evaluate the polynomial equality at \(t^2\). After multiplying by the state scale, the factors \((1-t^2)^N\) and \((1-t^2)^{-N}\) cancel, since \(1-t^2\ne0\). Thus \[ D_x(t,\varnothing) =t^{\ell(x)-\ell(b)}P_{x,b}(t^2) =p_{x,b}(t). \]
Step 3. Compute the full-kernel Hilbert value
Work now on \([1,b]\) with the stated section duality. Put \(d=\ell(b)-\ell(x)\). Its pairing shift is \(\ell(b)\), and there are \(\ell(x)\) incoming edges at \(x\). Stalk–costalk reciprocity gives \[ H_{L_x(U_x)}(t^2) =(t^2)^d C_{x,b}(t^{-2})(1-t^2)^{-N}. \] Use the character equality at \(t^{-2}\) to replace \(C_{x,b}\) by \(P_{x,b}\). This is a finite polynomial evaluation; the Hilbert sum itself remains at \(t^2\).
Step 4. Cancel the full-kernel scale
The coefficient factors again cancel. Since \(t\ne0\), \[ t^{-d}(t^2)^d=t^d=(t^{-1})^{-d}. \] Consequently \[ D_x(t,U_x) =t^dP_{x,b}(t^{-2}) =(t^{-1})^{-d}P_{x,b}((t^{-1})^2) =p_{x,b}(t^{-1}). \] This proves the second formula with precisely the supplied duality and character hypotheses.
References
The local source declarations below give the construction and proof.
Lean formalization
OAI.KLInvariance.BruhatGraph.normalized_emptyKernel_of_character: PositiveTransfer.lean:693.
OAI.KLInvariance.BruhatGraph.normalized_fullKernel_of_duality_character: PositiveTransfer.lean:708.