Alphabeta Math
Pipeline-generated
How statement and proof provenance work

The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.

  • Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
  • AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
  • AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.

These labels describe origin, not correctness: citations and verification chips remain separate evidence.

✓ 5 results · all verified · 0 also independently AI-judged
Every result on this page is machine-checked by a proof checker and read in full and owner-audited; the judge is an additional, independent cross-model AI review of the proofs. The 5 not AI-judged were verified by owner audit (typically over a confirmed judge false positive), not failures.

Soergel Intersection Forms and Hodge Theory — Examples

1 · Prerequisites

2 · Summary

This companion collects finite computations for soergel-intersection-forms-and-hodge-theory. The rank-one example computes the trace form, the reduced module and the Lefschetz map of Bs for n=2; the rank-two examples compute the character of BsBt and the braid decomposition BsBtBs≅Bsts⊕Bs, and identify the coinvariant algebra of S3 with the longest-element bimodule, including its Kazhdan-Lusztig expansion.

The final example works the hard Lefschetz and Hodge-Riemann signs of the coinvariant algebra of S3 in an explicit two-variable model, with the normalized top trace Tr⁡(x2y)=1, and then verifies the same signs for every element of the dominant chamber.

3 · Logical flowchart

4 · Definitions, theorems and proofs

ExampleConstruction: Literature-sourcedVerification: AI-adaptedprecheck passOpen item page →

The rank-two Bott-Samelson bimodule B_sB_t: character, decomposition and the braid relation

Example

Take n=3 and the two simple reflections s=s1, t=s2. The rank-two Bott-Samelson bimodule BsBt has character ch⁡(BsBt)=ch⁡(Bs)ch⁡(Bt)=(Hs+v)(Ht+v)=Hst+vHs+vHt+v2 in the standard basis, so its Δ-multiplicities are 1 at Hst, 1 at Hs and at Ht with shift v, and 1 at the identity with shift v2; in particular all multiplicities are non-negative and dim⁡(BsBt‾)=v2+2+v−2, whose value at v=1 is the ordinary real dimension 4. In S3 the product st has length 2 and Bst has the same character H‾st=Hst+vHs+vHt+v2, so BsBt=Bst is already indecomposable; equivalently the only summand is the top one Bst and all multiplicities my for y<st vanish. Similarly BsBtBs=Bsts⊕Bs with characters H‾sts+H‾s, whereas BtBsBt=Bsts⊕Bt; the two Bott-Samelson products are not isomorphic; the example also verifies the rank-two values of the Kazhdan-Lusztig polynomials Pe,st=Ps,st=Pt,st=1 used on the companion page.

Verification

technique · direct; multiplicativity of the character, the explicit two-letter Hecke computation, and comparison with the rank-two decomposition

Given: The simple reflections s=s1, t=s2 of S3, the one-letter bimodules Bs,Bt and the Bott-Samelson product BsBt (The Bott–Samelson bimodule of a word, The real type-A Soergel category by scalar extension), the character with its Δ-multiplicities (Standard graph bimodules, support filtrations and characters, The type-A standard character is multiplicative), the normalized Hecke algebra with its two bases and H‾s=Hs+v (Existence and uniqueness of the Kazhdan–Lusztig basis), the rank-two decomposition theorem (Rank-two type-A Soergel bimodule decompositions).

[F1] The character is multiplicative and satisfies ch⁡(Bsi)=Hsi+v; the coefficient of Hy in ch⁡(M) is ∑d(M:Δy(d))vd, and the reduced dimension of a layer Δy(d) is one in degree ℓ(y)−d (The type-A standard character is multiplicative, Standard graph bimodules, support filtrations and characters).

[F2] The standard basis of the Hecke algebra multiplies by HuHs=Hus when ℓ(us)=ℓ(u)+1; the Kazhdan-Lusztig basis is the unique basis H‾w∈Hw+∑y<wvZ[v]Hy of bar-invariant elements (Existence and uniqueness of the Kazhdan–Lusztig basis).

[F3] The rank-two decomposition theorem: BsBtBs≅Bsts⊕Bs with no shifts; exchanging s and t gives BtBsBt≅Bsts⊕Bt (Rank-two type-A Soergel bimodule decompositions).

[F4] The polynomial ring has only nonnegative degrees and degree-zero part R. The generator has left basis 1⊗1,1⊗xi for si, since S=Ssi⊕xiSsi; tensor products have the corresponding finite left bases (The real type-A Soergel category by scalar extension).

1.1givenF1F2

The character of the two-letter product. By [F1] and multiplicativity, ch⁡(BsBt)=(Hs+v)(Ht+v)=HsHt+vHs+vHt+v2; since st is reduced, [F2] gives HsHt=Hst, so the standard-basis expansion is Hst+vHs+vHt+v2. Its coefficients give the multiplicities 1,1,1,1 displayed, and reading the layers through [F1] the reduced degrees are 2,0,0,−2, so the graded dimension is v2+2+v−2 and the ordinary real dimension is 4.

2.1step 1.1F2

The Kazhdan-Lusztig element of st. Put A:=v−v−1 and J:=Hst+v(Hs+Ht)+v2; then J∈Hst+vZ[v]Hs+vZ[v]Ht+v2Z[v]. Since Hs−1=Hs+A, Ht−1=Ht+A and Hst‾=Hts−1=Hs−1Ht−1=(Hs+A)(Ht+A)=Hst+A(Hs+Ht)+A2 (the last identity uses HsHt=Hst), applying the semilinear involution gives J‾=Hst‾+v−1(Hs‾+Ht‾)+v−2=Hst+A(Hs+Ht)+A2+v−1(Hs+Ht)+2v−1A+v−2, whose coefficients are A+v−1=v at Hs+Ht and A2+2v−1A+v−2=(v2−2+v−2)+(2−2v−2)+v−2=v2 at He; hence J‾=J, and by uniqueness [F2], J=H‾st.

3.1step 1.1step 2.1F4algebra

Indecomposability of BsBt. The unshifted tensor M=S⊗SsS⊗StS is generated as a bimodule by 1⊗1⊗1. To see this, write a middle polynomial as h0+x1h1 with h0,h1∈Ss, which slide to the first slot. The remaining x1 slides to the last slot because t=s2 fixes x1. Therefore every tensor is a sum of p(1⊗1⊗1)q. After the shift (2) its lowest component is one-dimensional in degree −2, by its finite left tensor basis and S0=R. A degree-zero bimodule endomorphism must send the generator to a scalar multiple of itself, and generation forces that same scalar on all of M. Hence its degree-zero endomorphism algebra is R, with no nontrivial idempotent, so BsBt is indecomposable. The balanced map p⊗q⊗r↦p s(q) st(r) is surjective onto the st graph bimodule, showing that this indecomposable has the top graph of its reduced word; it is the normalized summand Bst. Thus BsBt≅Bst with no lower summands, independently of the Hom formula.

4.1step 2.1step 3.1F3∎

The three-letter braid relation. By multiplicativity and the products computed in steps 1.1 and 2.1, ch⁡(BsBtBs)=H‾sH‾tH‾s=H‾stH‾s; expanding with the Hecke relations gives H‾stH‾s=H‾sts+H‾s Indeed the expansion is J0+Hs+v where J0=Hsts+v(Hst+Hts)+v2(Hs+Ht)+v3. The element J0=H‾stH‾s−H‾s is bar fixed, with triangular lower coefficients in vZ[v], so uniqueness in [F2] identifies it with H‾sts. The ordinary real dimensions of the reductions add to 6+2=8. Exchanging s and t in [F3] gives BtBsBt≅Bsts⊕Bt, whose character differs by H‾t−H‾s=Ht−Hs≠0, so these Bott-Samelson products are not isomorphic; the rank-two Kazhdan-Lusztig data Pe,st=Ps,st=Pt,st=1 are the coefficients displayed in step 2.1. This verifies the example.

Remarks

  • The reduction to rank two is essential here: the computation uses the explicit two- and three-letter decompositions and is an illustration only; the design of the page forbids its use as a supplier inside the general induction, whose base is computed on the axiom page.
  • The two-letter indecomposability is established by its scalar degree-zero endomorphisms; the triple decomposition is the explicit rank-two theorem. Neither computation uses the general character theorem.
ExampleConstruction: Literature-sourcedVerification: AI-adaptedprecheck passOpen item page →

The coinvariant algebra of S_3 as the Soergel bimodule of the longest element

Example

Take n=3 and let w0=s1s2s1 be the longest element of S3. Then Bw0=S⊗SS3S(3) and its reduced module is C(3), where C=S/(S+S3) is the unshifted coinvariant algebra: C has the Poincare polynomial 1+2t2+2t4+t6, so dim⁡C=6, and in the centered grading dim⁡(Bw0‾)−3=dim⁡(Bw0‾)3=1,dim⁡(Bw0‾)−1=dim⁡(Bw0‾)1=2. The standard-basis expansion of the Kazhdan-Lusztig basis element is H‾w0=Hw0+v(Hs1s2+Hs2s1)+v2(Hs1+Hs2)+v3, obtained by solving the bar-invariance and triangularity conditions with coefficients in vdZ[v−2]; its coefficients are non-negative, its value at v=1 sums to 6=dim⁡C, and it agrees with the rank-one and rank-two Kazhdan-Lusztig data Pe,w0=Psi,w0=Pw,w0=1 for ℓ(w)=2. Hence ch⁡(Bw0)=H‾w0: Soergel's conjecture holds for the longest element of S3, and the Δ-multiplicities of the coinvariant bimodule are the six coefficients just displayed, one for each element of S3.

Verification

technique · direct; the identification with the invariants-tensor bimodule, an explicit bar-invariance computation in the Hecke algebra, and the comparison of the graded dimensions

Given: S=R[x1,x2,x3] with deg⁡xi=2, the longest element w0=s1s2s1, the rank-two longest bimodule Bi,i+1,i=R⊗RWi,i+1R(3) (The rank-two longest type-A Soergel bimodule), the reduced module and the character (The reduced Soergel module and the dominant Lefschetz operator, Standard graph bimodules, support filtrations and characters), the normalized Hecke algebra with Hs2=1+(v−1−v)Hs and the Kazhdan-Lusztig basis (Existence and uniqueness of the Kazhdan–Lusztig basis), and multiplicativity of the standard character (The type-A standard character is multiplicative).

[F1] The rank-two longest bimodule is S⊗SS3S(3) after real scalar extension, and right reduction is tensoring with S/S+=R (The rank-two longest type-A Soergel bimodule, The reduced Soergel module and the dominant Lefschetz operator). The symmetric-polynomial and quotient-basis computations needed here are supplied in step 1.1; no general invariant-ring theorem is assumed.

[F2] In the Hecke algebra Hs−1=Hs+(v−v−1) for every simple s, and the bar involution acts on the standard basis by Hw‾=Hw−1−1; the Kazhdan-Lusztig basis is the unique basis of elements H‾w=Hw+∑y<wpy,wHy with py,w∈vZ[v] and H‾w‾=H‾w (Existence and uniqueness of the Kazhdan–Lusztig basis).

[F3] The coefficient of Hy in ch⁡(M) is ∑d(M:Δy(d))vd with non-negative integral multiplicities, and for the standard layer Δy(d) the reduced dimension sits in degree ℓ(y)−d; hence the multiset of monomials of the character, read with the degrees ℓ(y)−d, is the graded dimension of the reduced module (Standard graph bimodules, support filtrations and characters).

[F4] The character is multiplicative with simple value Hs+v, and the rank-two decomposition is Bs1Bs2Bs1≅Bw0⊕Bs1 (The type-A standard character is multiplicative, Rank-two type-A Soergel bimodule decompositions).

1.1givenF1

The bimodule and its reduction. Symmetric polynomials in three variables are generated by e1,e2,e3: for a homogeneous symmetric polynomial, its lexicographically leading monomial has exponents a≥b≥c; subtracting its coefficient times e1a−be2b−ce3c lowers that leading monomial. The finite set of monomials of the fixed total degree makes this reduction terminate. Thus SS3=R[e1,e2,e3] and [F1] gives the reduced module C(3) with C=S/(e1,e2,e3). Eliminating x3=−x−y gives C=R[x,y]/(x3,y2+xy+x2): the e3 relation is x2y+xy2=0, and subtracting this from x(x2+xy+y2)=0 gives x3=0, with the same identities proving the reverse inclusion of ideals. Over R[x]/(x3), division by the monic quadratic in y gives the free basis 1,y; hence 1,x,y,x2,xy,x2y is a real basis and the dimensions are 1,2,2,1 in degrees 0,2,4,6, centered at −3,−1,1,3 after shifting by (3). The tensor bimodule is generated as a bimodule by 1⊗1 and its bottom component is one dimensional, so every degree-zero endomorphism is scalar; it is therefore indecomposable. The balanced surjection f⊗g↦fw0(g) onto Sw0 proves that it has the w0 graph in its support. In the rank-two decomposition it is the summand carrying the w0 graph (the other summand Bs1 has only the e,s1 graphs), so its label is w0.

1.2givenF2algebra

The Kazhdan-Lusztig element. Put A:=v−v−1 and J:=Hw0+v(Hs1s2+Hs2s1)+v2(Hs1+Hs2)+v3. The element J lies in Hw0+∑y<w0vZ[v]Hy, since all coefficients belong to vZ[v]. For bar-invariance we expand: from Hs−1=Hs+A and the reduced products Hs1Hs2=Hs1s2 and Hs1s2Hs1=Hw0 one computes Hs1s2‾=Hs1s2+A(Hs1+Hs2)+A2 and Hw0‾=Hw0+A(Hs1s2+Hs2s1)+A2(Hs1+Hs2)+A+A3, while Hs1‾=Hs1+A and Hs2‾=Hs2+A. Applying the semilinear involution to J gives J‾=Hw0‾+v−1(Hs1s2‾+Hs2s1‾)+v−2(Hs1‾+Hs2‾)+v−3; collecting coefficients in the standard basis, the coefficient of Hs1s2+Hs2s1 becomes A+v−1=v, the coefficient of Hs1+Hs2 becomes A2+2v−1A+v−2=v2, and the coefficient of He becomes A+A3+2v−1A2+2v−2A+v−3=v3. Hence J‾=J, and by the uniqueness clause of [F2] J=H‾w0.

2.1step 1.1step 1.2F3F4∎

Character and multiplicities. By [F4], ch⁡(Bw0)=(Hs1+v)(Hs2+v)(Hs1+v)−(Hs1+v)=J, the last equality following by the quadratic relation Hs2=1+(v−1−v)Hs and reduced multiplication. Step 1.2 gives J=H‾w0, so the Δ-multiplicities of Bw0 are exactly the six coefficients of J, one for each of the six elements of S3. Reading them through [F3], the layer Δw0(0) contributes in degree ℓ(w0)=3, the two layers Δs1s2(1),Δs2s1(1) contribute in degree 1 with multiplicity two, the two layers Δs1(2),Δs2(2) contribute in degree −1 with multiplicity two and the layer Δe(3) contributes in degree −3; this is the graded dimension 1,2,2,1 of the centered coinvariant algebra computed in step 1.1, and the value at v=1 is 6=dim⁡C. The coefficients Pe,w0=Psi,w0=Ps1s2,w0=Ps2s1,w0=1 are the classical normalization of these coefficients. This verifies the example.

Remarks

  • The two computations are independent: the bar-invariance determines the Kazhdan-Lusztig element from the Hecke algebra alone, and the graded dimensions of the coinvariant algebra come from the invariant theory; their agreement is the content of Soergel's conjecture at w0.
  • The example is an illustration of the character theorem, not a supplier of it; the base of the simultaneous induction is computed on the axiom page and does not use this page.
CounterexampleConstruction: Literature-sourcedVerification: AI-adaptedprecheck passOpen item page →

Hard Lefschetz fails for a non-dominant degree-two element: the rank-one case

Statement refuted

In the rank-one case n=2 the element ρ′:=x1+x2 is a degree-two element of S with ρ′(αs∨)=0, so it is not strictly dominant. Its action on the reduced rank-one Soergel module is zero: ρ′=e1=x1+x2 is s-invariant and of positive degree, and x1+x2=1⋅e1 acts on the right basis element cid‾ as the right action by an element of S+, hence ρ′⋅cid‾=0; a fortiori it cannot map (Bs‾)−1=R cid‾ isomorphically to (Bs‾)1. Therefore for ρ′ the hard Lefschetz property fails (the map on the −1 degree is the zero map), even though ρ′ is a symmetric degree-two endomorphism of the reduced module; the same failure occurs for every degree-two element with ρ′(αs∨)=0. Consequently the hard Lefschetz and Hodge-Riemann conclusions do not hold for arbitrary degree-two elements when strict dominance is omitted. This does not assert that every Lefschetz element is strictly dominant.

Facts & Assumptions

Given: The rank-one setting n=2, S=R[x1,x2] with deg⁡xi=2, the simple reflection s=s1 with coordinate root α=αs=x1−x2, the generator Bs=S⊗SsS(1) with right basis cid=1⊗1 of degree −1 and cs=12(α⊗1+1⊗α) of degree +1, its reduced module Bs‾=Bs/BsS+ with the induced basis cid‾,cs‾, and the degree-two element ρ′=x1+x2 (The Soergel bimodule Bi of a simple reflection, The real type-A Soergel category by scalar extension, The reduced Soergel module and the dominant Lefschetz operator).

[F1]

The left and right actions on Bs satisfy r⋅cid=cid⋅sr+∂s(r)cs and r⋅cs=cs⋅r, and cscs=csα, cid is the unit of the commutative ring Bs(−1) (The trace, the invariant intersection form on Bott-Samelson bimodules and its reduction, The Soergel bimodule Bi of a simple reflection).

[F2]

The reduction functor kills right multiplication by S+: M‾=M/MS+, and mr+MS+=MS+ for r∈S+ (The reduced Soergel module and the dominant Lefschetz operator).

[F3]

Hard Lefschetz for a degree-two symmetric operator L on a graded space with lowest degree −d requires L:( )−1→( )1 to be an isomorphism in the case d=1; definiteness and symmetry conventions as in Primitive subspaces, hard Lefschetz and the Hodge-Riemann signs: the dictionary of hypotheses S, hL and HR and Bilinear forms, and symmetric, skew-symmetric, and alternating bilinear forms.

Counterexample

technique · direct computation in the rank-one reduced module
1.1givenF1

Non-dominance. The reflection s exchanges x1 and x2, so ρ′=x1+x2 is s-invariant and ρ′(α∨)=(x1+x2,x1−x2)=1−1=0 in the standard form of the coordinate normalization; in particular ρ′ is not strictly dominant.

2.1step 1.1givenF1

The left action of ρ′ equals its right action. Since ρ′∈Ss, the balanced tensor relation gives ρ′⊗1=1⊗ρ′ in Bs; hence for every b∈Bs the left multiplication ρ′⋅b equals the right multiplication b⋅ρ′: indeed both actions are S-linear and the identity holds on the two basis elements cid (where ρ′⋅cid=ρ′⊗1=1⊗ρ′=cid⋅ρ′) and cs (where both equal csρ′ by [F1]).

3.1step 2.1F1F2

The operator vanishes on the reduced module. By step 2.1, ρ′⋅cid=cid⋅ρ′=1⊗ρ′ lies in BsS+, since it is the right product of cid=1⊗1 with ρ′∈S+; by [F2] its class in Bs‾ is zero, so ρ′⋅cid‾=0. Similarly ρ′⋅cs=cs⋅ρ′=csρ′ lies in BsS+ and reduces to zero. Since cid‾ and cs‾ span Bs‾, the induced operator L=ρ′⋅(−) is the zero endomorphism of Bs‾.

4.1step 3.1F3

Failure of hard Lefschetz. The reduced module has dim⁡(Bs‾)−1=dim⁡(Bs‾)1=1 with bases cid‾ and cs‾. The degree-one Lefschetz map L:(Bs‾)−1→(Bs‾)1 is zero by step 3.1, hence not an isomorphism, and hard Lefschetz fails for L in the sense of [F3]. The bottom Lefschetz form ⟨−,L(−)⟩ is likewise identically zero, so the Hodge-Riemann relations with any nonzero sign also fail.

5.1step 4.1∎

The general degree-two case. Let ρ′′=ax1+bx2 satisfy ρ′′(α∨)=a−b=0; then a=b and ρ′′=aρ′ is a scalar multiple of ρ′. If a=0 the operator is trivially zero; if a≠0 the same computation as in steps 2.1-4.1 applies because invariance and positivity of degree are unaffected by the scalar. Hence every degree-two element with vanishing evaluation on αs∨ yields the zero operator on the reduced rank-one module and fails hard Lefschetz. This proves the claimed failure and shows that allowing arbitrary degree-two elements invalidates the conclusion; no choice principle is used.

Remark

The counterexample isolates the hypothesis at the very bottom of the induction: the strict dominance of ρ is used on this page exactly through the positivity of the scaling factors (wρ)(αs∨) for reduced prefixes, and at the simple reflection itself the vanishing of the coroot evaluation makes the operator vanish on the whole reduced module. Since ρ′ is still symmetric of degree two, the failure is not an artifact of a badly normalized form.

ExampleConstruction: Literature-sourcedVerification: AI-adaptedprecheck passOpen item page →

The rank-one Soergel bimodule: trace form, reduced module and hard Lefschetz

Example

Take n=2, so Sn={e,s} and Bs=BsR=S⊗SsS(1) with S=R[x1,x2], deg⁡xi=2 and αs=x1−x2. In the right S-basis {cid,cs} of degrees −1 and +1 the multiplication relations are cidcid=cid,cidcs=cs,cscs=csαs. The trace is Tr⁡(cida+csb)=b for a,b∈S and the intersection form has the Gram matrix (⟨cid,cid⟩⟨cid,cs⟩⟨cs,cid⟩⟨cs,cs⟩)=(011αs), whose determinant is −1, so the form is non-degenerate, and the induced form on the reduced module is non-degenerate too. The reduced module has the real basis (cid‾,cs‾) with degrees −1 and +1, so dim⁡(Bs‾)−1=dim⁡(Bs‾)1=1; for ρ=αs/2=(x1−x2)/2, which is strictly dominant with ρ(αs∨)=1, one computes ρ⋅cid‾=cs‾ and ρ⋅cs‾=0. Hence ρ:(Bs‾)−1→(Bs‾)1 is an isomorphism, hard Lefschetz holds, the primitive subspace in degree −1 is all of (Bs‾)−1, and the Lefschetz form (a,b)=⟨a,ρb⟩ is positive on it: ⟨cid‾,cs‾⟩=1>0, matching the Hodge-Riemann sign (−1)(ℓ(s)−1)/2=+1. Finally ch⁡(Bs)=Hs+v=H‾s, the Kazhdan-Lusztig basis element of the simple reflection, which is Soergel's conjecture at s, and the rational model over Q gives the same numbers.

Verification

technique · direct computation in the right basis of the rank-one bimodule

Given: The rank-one data above: Bs=S⊗SsS(1) with right basis cid=1⊗1 (degree −1) and cs=12(αs⊗1+1⊗αs) (degree +1), the trace and intersection form of the Bott-Samelson bimodule (The trace, the invariant intersection form on Bott-Samelson bimodules and its reduction, The Soergel bimodule Bi of a simple reflection), the reduced module and dominant operator (The reduced Soergel module and the dominant Lefschetz operator, The real type-A Soergel category by scalar extension), and the HR dictionary (Primitive subspaces, hard Lefschetz and the Hodge-Riemann signs: the dictionary of hypotheses S, hL and HR).

[F1] The left action satisfies r⋅cid=cid⋅s(r)+∂s(r)cs and r⋅cs=cs⋅r; the products are cidcid=cid, cidcs=cs, cscs=csαs; the trace is the coefficient of cs in the right basis, and ⟨b,b′⟩=Tr⁡(bb′) (The trace, the invariant intersection form on Bott-Samelson bimodules and its reduction).

[F2] The reduction M‾=M/MS+ kills right multiples by positive-degree elements; a non-degenerate invariant form descends to a non-degenerate form on the reduction, and the induced form on the reduced module is the one used in the HR dictionary (The reduced Soergel module and the dominant Lefschetz operator, The trace, the invariant intersection form on Bott-Samelson bimodules and its reduction).

[F3] For a graded space with lowest degree −d hard Lefschetz requires ρi:( )−i→( )i to be isomorphisms, the primitive subspaces are P−i=ker⁡(ρi+1) and the standard sign in degree −i is (−1)(d−i)/2 (Primitive subspaces, hard Lefschetz and the Hodge-Riemann signs: the dictionary of hypotheses S, hL and HR, Hard Lefschetz gives the primitive decomposition, orthogonality and the signature formula).

[F4] The standard character sums the Δy(d) layers with coefficients vdHy, where Δy(d)=Sy(d−ℓ(y)) (Standard graph bimodules, support filtrations and characters). The bar-fixed triangular element at a simple reflection is Hs+v=H‾s (Existence and uniqueness of the Kazhdan–Lusztig basis).

1.1givenF1

Products and the trace. The multiplication rules are those of [F1]. By definition the trace extracts the coefficient of cs in the right basis, so Tr⁡(cida+csb)=b for a,b∈S.

2.1step 1.1F1F2

The Gram matrix and non-degeneracy. Using ⟨b,b′⟩=Tr⁡(bb′) and the products of step 1.1, ⟨cid,cid⟩=Tr⁡(cid)=0, ⟨cid,cs⟩=Tr⁡(cs)=1 and ⟨cs,cs⟩=Tr⁡(csαs)=αs, so the Gram matrix in the basis (cid,cs) is (011αs); its determinant is −1, hence the form is non-degenerate, and by [F2] so is the induced form on the reduced module.

2.2step 1.1F1F2

The reduced action of ρ. Let ρ=αs/2. Since ρ is s-anti-invariant, s(ρ)=−ρ and ∂s(ρ)=1, so by [F1] we get ρ⋅cid=cid⋅(−ρ)+cs=cs−cidρ; reducing kills the right multiple cidρ by [F2], so ρ⋅cid‾=cs‾. Similarly ρ⋅cs=csρ by [F1] and csρ is a right multiple by a positive-degree element, so ρ⋅cs‾=0. Also ρ(αs∨)=1>0, so ρ is strictly dominant.

3.1step 2.1step 2.2F3

Hard Lefschetz, primitivity and the sign. The reduced module has real basis cid‾ in degree −1 and cs‾ in degree +1, hence both graded pieces are one-dimensional; by step 2.2 the degree-one Lefschetz map ρ:(Bs‾)−1→(Bs‾)1 sends a basis vector to a basis vector and is an isomorphism, which is hard Lefschetz by [F3]. The primitive subspace is P−1=ker⁡(ρ2∣(Bs‾)−1)=(Bs‾)−1, since ρ2 maps the one-dimensional space to the higher space in two steps and ρ2 lands in degree +3, which is zero; the Lefschetz form on it is ⟨cid‾,ρcid‾⟩=⟨cid‾,cs‾⟩=1>0, the reduction of the Gram entry 1, and the standard sign in degree −1 is (−1)(1−1)/2=+1, so HR holds with the standard sign and the bottom form is positive.

4.1step 3.1F1F2F4∎

The character. Multiplication Bs→S(1) is surjective with kernel generated on the right by k=cs−cidαs, of degree 1. The left-action rule gives rk=k s(r), so the kernel is Ss(−1)=Δs(0) and the quotient is S(1)=Δe(1). These two layers give ch⁡(Bs)=Hs+v=H‾s. The top graph layer has generator degree +1 and the unit quotient has generator degree −1. The rational model has the same formulas after real scalar extension.

Remarks

  • This is the base case of the rank-one computations used both in the counterexample on the non-dominant element and in the base of the simultaneous induction; the two-dimensional reduced module has a one-dimensional primitive space in degree −1, and the sign +1 is the standard sign.
  • The trace normalization is the one of the Bott-Samelson intersection form; no freedom remains after the normalizations of the page, since the form is unique up to scalar and the trace fixes the scalar by Tr⁡(cs)=1.
ExampleConstruction: Literature-sourcedVerification: AI-adaptedprecheck passOpen item page →

Hard Lefschetz and the Hodge-Riemann signs for the coinvariant algebra of S_3

Example

For n=3 identify the centered reduced longest-element module with C(3), where C=R[x1,x2,x3]/(e1,e2,e3) has dimensions 1,2,2,1 in degrees 0,2,4,6, so that the centered degrees are −3,−1,1,3. For every strictly dominant ρ hard Lefschetz is the pair of isomorphisms ρ3:C(3)−3→C(3)3 and ρ:C(3)−1→C(3)1. The primitive spaces used in the Hodge-Riemann relations are P−3=C(3)−3 of dimension one and P−1=ker⁡(ρ2:C(3)−1→C(3)3) of dimension one; their Lefschetz forms are respectively positive and negative definite, and the primitive decomposition consists of one length-four ρ-string and one length-two ρ-string. For an explicit strictly dominant choice let x=x1, y=x2, x3=−x−y, ρ=x1−x3=2x+y and normalize Tr⁡(x2y)=1. Then C=R[x,y]/(x2+xy+y2,x3), with basis 1;x,y;x2,xy;x2y; the ρ-map from degree 2 to degree 4 has the matrix (2−111), ρ3(1)=6x2y, P−1=Ry and Tr⁡(yρy)=−2. These computations verify hard Lefschetz and the primitive signs directly in rank two.

Verification

technique · direct; elimination of the linear invariant, reduction to a two-variable model, and explicit matrix and trace computations in both the special and the general dominant chamber

Given: The coinvariant algebra C=R[x1,x2,x3]/(e1,e2,e3) and the identification with the centered reduced module of Bw0 (The coinvariant algebra of S_3 as the Soergel bimodule of the longest element), the operator ρ⋅(−) of left multiplication by a strictly dominant degree-two element and the trace normalization (The reduced Soergel module and the dominant Lefschetz operator, The trace, the invariant intersection form on Bott-Samelson bimodules and its reduction), and the hard-Lefschetz and Hodge-Riemann dictionary with the sign (−1)(d−i)/2 for lowest degree −d (Primitive subspaces, hard Lefschetz and the Hodge-Riemann signs: the dictionary of hypotheses S, hL and HR, Hard Lefschetz gives the primitive decomposition, orthogonality and the signature formula).

[F1] The centered module C(3) has lowest degree −3, graded dimensions 1,2,2,1 in degrees −3,−1,1,3 and a non-degenerate invariant form normalized by the positive bottom value; for it, hard Lefschetz is ρ3:C−3→C3 together with ρ:C−1→C1, and the standard sign in degree −i is (−1)(3−i)/2, so the primitive space P−1⊂C−1 must be negative definite and P−3=C−3 positive definite (The coinvariant algebra of S_3 as the Soergel bimodule of the longest element, Primitive subspaces, hard Lefschetz and the Hodge-Riemann signs: the dictionary of hypotheses S, hL and HR).

[F2] In the quotient the relations e1=e2=e3=0 hold and, after the substitution x3=−x−y, are equivalent to x2+xy+y2=0 and x3=0; the reduction of the intersection form is the standard top trace with Tr⁡(x2y)=1, and C is spanned by the six monomials of the display (The coinvariant algebra of S_3 as the Soergel bimodule of the longest element; the trace identification is checked in step 2.1).

1.1givenF2

The two-variable model. Put x=x1, y=x2 and x3=−x−y. Then e1=x+y+x3=0, while e2=x1x2+x1x3+x2x3=xy+x(−x−y)+y(−x−y)=−(x2+xy+y2) and e3=x1x2x3=−x2y−xy2; hence the coinvariant ideal is generated by x2+xy+y2 and x2y+xy2, and since x3=x(x2+xy+y2)−x2y−xy2 the second relation can be replaced by x3=0, giving C=R[x,y]/(x2+xy+y2,x3). The monomials 1;x,y;x2,xy;x2y span it: the relations x2=−xy−y2 and x3=0, together with xy2=−x2y and y3=0, reduce every monomial of degree ≥2 in x to one of them, and their independence follows by monic division by y2+xy+x2 over R[x]/(x3), which gives the free basis 1,y over that ring.

2.1step 1.1givenF2algebra

The trace and pairing. The rank-two summand map j(p⊗q)=p⊗1⊗1⊗q pulls the Bott–Samelson trace back to ∂1∂2∂1(p)q, as follows by taking the right-coefficient trace successively in the three slots; this is the explicit trace calculation in the rank-two decomposition proof, also obtained directly from the rank-one trace convention. After right reduction it is the top trace on C, since ∂1(x12x2)=x1x2, ∂2(x1x2)=x1 and ∂1(x1)=1. Hence Tr⁡(x2y)=1 and all degrees other than 6 have zero trace. The complementary-degree pairings are 1 in degrees 0,6 and (011−1) in degrees 2,4, of determinant −1; they prove nondegeneracy directly. The form used below is therefore the actual induced reduced intersection form, without invoking the general Hodge theorem.

3.1step 1.1step 2.1

The special dominant element. For ρ=2x+y one computes ρx=2x2+xy and ρy=2xy+y2=−x2+xy using y2=−x2−xy; the matrix of ρ from the degree-2 block with basis (x,y) to the degree-4 block with basis (x2,xy) is therefore (2−111), whose determinant is 3≠0, so ρ:C−1→C1 is an isomorphism. Moreover ρ2x=3x2y and ρ2y=0: indeed x2ρ=x2(2x+y)=2x3+x2y=x2y, xyρ=xy(2x+y)=2x2y+xy2=x2y; hence ρ2x=2⋅x2y+x2y=3x2y and ρ2y=−x2y+x2y=0; hence P−1=ker⁡(ρ2)=Ry and ρ3(1)=ρ2(2x+y)=6x2y≠0, so ρ3:C−3→C3 is an isomorphism as both sides are one-dimensional.

4.1step 3.1F1

The signs for the special element. With the normalization Tr⁡(x2y)=1, the bottom primitive form is (1,1)=Tr⁡(1⋅ρ3(1))=Tr⁡(6x2y)=6>0, positive as required in degree −3; the middle primitive form is (y,y)=Tr⁡(y⋅ρy)=Tr⁡(y(−x2+xy))=Tr⁡(−x2y+xy2)=−1−1=−2<0, negative as required in degree −1, where the standard sign is (−1)(3−1)/2=−1. The primitive decomposition is read off from the two strings: 1↦2x+y↦3x2+3xy↦6x2y spans the four degrees with one vector each, and y↦−x2+xy spans degrees −1 and +1; together they account for all four graded pieces and their dimensions, so there are exactly one length-four string and one length-two string.

5.1step 4.1F1

The general dominant chamber. Let ρ=c1x1+c2x2+c3x3 be strictly dominant, so c1>c2>c3; subtracting a multiple of e1 (which is zero in C) writes its class as ρ=ax+by with a=c1−c3>b=c2−c3>0. Then ρx=ax2+bxy and ρy=−bx2+(a−b)xy, so the middle matrix is (a−bba−b) with determinant D=a2−ab+b2>0, and ρ3(1)=3ab(a−b)x2y: indeed ρ2x=b(2a−b)x2y and ρ2y=a(a−2b)x2y, so p:=a(a−2b)x−b(2a−b)y satisfies ρ2p=0, and expanding gives Tr⁡(pρp)=A2b−B2a+2AB(b−a) with A=a(a−2b), B=b(2a−b), which simplifies to −3ab(a−b)(a2−ab+b2)<0. Hence P−1=Rp (one-dimensional, since p≠0 for a>b>0) carries a negative definite Lefschetz form, the bottom form Tr⁡(3ab(a−b)x2y)>0 is positive, and every strictly dominant ρ satisfies hard Lefschetz with the standard signs on both primitive spaces.

6.1step 4.1step 5.1F1∎

Conclusion. Steps 2.1 and 3.1 verify the displayed isomorphisms, the two primitive spaces and their signs for the explicit choice, and step 5.1 shows the same for every strictly dominant element; the primitive decomposition has the two strings claimed. This verifies the example by its quotient and trace calculations.

Remarks

  • Define the top trace to be zero off degree 6 and to send x2y to 1. Its multiplication pairing is non-degenerate: the degree-0 and degree-6 pieces pair by 1, and the degree-2 and degree-4 pairing in the displayed bases has matrix (011−1), of determinant −1. The divided-difference normalization agrees, since ∂1(x12x2)=x1x2, ∂2(x1x2)=x1 and ∂1(x1)=1. This is a positive scalar normalization of the invariant form, as its bottom Lefschetz value is positive.
  • The example is an illustration in rank two and is not used as a supplier in the simultaneous induction; the general statement for w0 is proved on the axiom page by the induction itself.

5 · Examples, counterexamples and false statements

None yet.

Sources