Three homogeneous embeddings of \(\mathit{DW}(2n-1,2)\) (Q670212)

Given a point-line incidence geometry \(\Gamma=(P,L)\) and a vector space \(V\), a full projective embedding of \(\Gamma\) is an injective map \(\varepsilon:P\to{\mathrm{PG}}(V)\) such that the image of \(\varepsilon\) spans \(V\) and for each \(\ell\in L\) the set \(\varepsilon(\ell):=\{ \varepsilon(p) : p\in\ell\}\) is a line of \(\mathrm{PG}(V)\). The dimension of \(\varepsilon\) is defined as the vector dimension of \(V\). Projective embeddings provide insight on the geometry, as well as a useful ``concrete'' model for it. An embedding \(\varepsilon':P\to{\mathrm{PG}}(V')\) is said to be a \textit{quotient} of an embedding \(\varepsilon\) when there is a subspace \(W\leq V\) such that \(W\) is disjoint from the image of each line of \(\Gamma\), \(V'=V/W\) and \(\varepsilon'(p)=\langle\varepsilon(p),W\rangle\) for all \(p\in P\). A projective embedding \(\varepsilon\) of \(\Gamma\) is \textit{universal} if any projective embedding of the geometry is isomorphic to a quotient of \(\varepsilon\). In a sense, universal embeddings of \(\Gamma\) are the largest possible full projective embeddings of the geometry. It is known that the vector dimension of the universal embedding of the dual polar space \(DW(2n-1,2)\) must be \(\frac{1}{3}(2^n+1)(2^{n-1}+1)\); the structure of this embedding, however, is not well understood. In this very nice paper, the author constructs a homogeneous full projective embedding \(\varepsilon_h\) of the dual polar space \(DW(2n-1,2)\) from hyperplane intersections of hyperbolic type of the associated quadric \(Q(2n,2)\). This embedding is conjectured to be universal; as a byproduct of this investigation, two new homogeneous full projective embeddings \(\varepsilon_1\) and \(\varepsilon_2\) of \(DW(2n-1,2)\) with \(n\geq 3\) of dimension respectively \(\frac{1}{3}(2^{n-1}-2)+2^{n-1}\) and \(\frac{1}{3}(2^{2n-1}-2)+2^{n-1}-2n\) are obtained as quotients.