Conjecture on irreducible components of varieties with symmetric compound matrices

About 2 years old · traced to

Let NN be the matrix-size parameter, let [N+1]={1,…,N+1}[N+1]=\{1,\ldots,N+1\}, and let S=S2CN+1S=S^2\mathbb{C}^{N+1} be the subspace of symmetric (N+1)×(N+1)(N+1)\times(N+1) matrices. For r∈[N+1]r\in[N+1], define

SCr={B∈C(N+1)×(N+1)∣Cr(B) is symmetric},\mathcal{SC}_r=\{B\in\mathbb{C}^{(N+1)\times(N+1)}\mid C_r(B)\text{ is symmetric}\},

where Cr(B)C_r(B) is the rrth compound matrix of BB, and write (Xj)Sj∨(X_j)_{S_j}^\vee for the relative dual variety introduced earlier. Let A=⋀2CN+1A=\bigwedge^2\mathbb{C}^{N+1} denote the space of skew-symmetric matrices. Irreducible-component conjecture. For every r∈[N+1]r\in[N+1]: (i) SC1=S=S2CN+1\mathcal{SC}_1=S=S^2\mathbb{C}^{N+1} is an irreducible component of SCr\mathcal{SC}_r; (ii) (Xn−r)Sn−r∨(X_{n-r})_{S_{n-r}}^\vee is an irreducible component of SCr\mathcal{SC}_r for every 2≤r≤N+12\le r\le N+1; and (iii) A=⋀2CN+1A=\bigwedge^2\mathbb{C}^{N+1} is an irreducible component of SCr\mathcal{SC}_r for all N≥3N\ge 3 and even r∈[N]r\in[N]. The conjecture is motivated by symbolic computations for N=2N=2 and N=3N=3, which exhibit several of these components, while the preceding example also shows that the varieties SCi\mathcal{SC}_i need not be nested. Determining all irreducible components in general remains open.

References

Primary source

Vahid Shahverdi, Giovanni Luca Marchetti and Kathlén Kohn, “On the Geometry and Optimization of Polynomial Convolutional Networks”, arXiv:2410.00722 (2026).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.