On \(\lambda \)-homomorphic skew braces (Q2069808): Difference between revisions

A skew brace is a triple \((A,+,\circ)\), where \((A,+)\) and \((A,\circ)\) are groups and the compatibility condition \(a\circ (b+c)=a\circ b-a+a\circ c\) holds for all \(a,b,c\in A\). If \(A\) is a skew brace, then \(\lambda\colon (A,\circ)\to\operatorname{Aut}(A,+)\), where \(\lambda_a(b)=-a+a\circ b\), is a group homomorphism. The paper concerns \(\lambda\)-homomorphic skew braces, i.e., skew braces where the map \(\lambda\colon (A,+)\to\operatorname{Aut}(A,+)\) is a group homomorphism, that is \(\lambda_{a+b}(c)=\lambda_a\lambda_b(c)\) for all \(a,b,c\in A\). The authors prove that \(\lambda\)-homomorphic skew braces are meta-trivial (Theorem 2.12). Note that a skew brace \(A\) is meta-trivial if and only if the ideal \(A*A\), defined as the additive subgroup of \(A\) generated by \(a*b=\lambda_a(b)-b\), is trivial as a skew brace, see Proposition 2.3 of [\textit{F. Cedó} et al., Proc. Lond. Math. Soc. (3) 118, No. 6, 1367--1392 (2019; Zbl 1432.16031)]. In Question 2.13, the authors ask if meta-trivial skew braces are \(\lambda\)-homomorphic. This is not true. Take for example \(A\) the symmetric group \(\mathbb{S}_3\) in three letters with operations \(a+b=ab\) and \(a\circ b=ba\). Then \(A\) is a skew brace that is meta-trivial and not \(\lambda\)-homomorphic. The authors also consider \(\lambda\)-cyclic skew braces. Those are skew braces where the image of \(\lambda\) is a cyclic subgroup of \(\operatorname{Aut}(A,+)\). In Theorem 5.8, it is proved that \(\lambda\)-cyclic skew braces are symmetric (i.e., \((A,\circ,+)\) is also a skew braces). Symmetric skew braces are known as bi-skew braces in the theory of Hopf-Galois extensions, see [\textit{L. N. Childs}, New York J. Math. 25, 574--588 (2019; Zbl 1441.12001)]. Exact group factorizations produce \(\lambda\)-cyclic skew braces (Theorem 5.10).