Involutive Heegaard Floer homology (Q2359429): Difference between revisions

In the paper at hand, the authors define a 3-manifold invariant of spin\(^c\) 3-manifolds called involutive Heegaard Floer homology, which they conjecture to be isomorphic to \(\mathbb{Z}/4\mathbb{Z}\)-equivariant Seiberg-Witten homology. The construction goes roughly as follows: given a pointed Heegaard diagram \(\mathcal{H} = (\Sigma,\alpha,\beta,z)\), there is a map \(\iota: CF(\mathcal{H} \to CF(\mathcal{H})\) induced by swapping the \(\alpha\)- and \(\beta\)-curves and reversing the orientation of \(\Sigma\) (and by a sequence of moves relating \((\Sigma,\alpha,\beta,z)\) and \((-\Sigma,\beta,\alpha,z)\)); the map \(\iota^2\) is known to be chain homotopic to the identity. The involutive Heegaard Floer homology is the homology of the mapping cone of \((1+\iota)\). The authors then turn to defining maps induced by cobordisms (with additional choices, due to current lack of sufficient naturality in ordinary Heegaard Floer homology), involutive correction terms, and prove some Frøyshov-type inequalities for spin cobordisms (Theorem 1.2). Finally, they prove a large surgery formula for involutive Floer homology, that allows them to give some concrete computations for (mirrors) of L-space knots (a class of knots that includes positive torus knots, and more generally links of irreducible curve singularities) and homologically thin knots (a class that includes alternating and quasi-alternating knots). As an application, they re-prove that the figure-eight knot is not slice; while this is classically known (e.g. because of the Fox-Milnor condition on the Alexander polynomial), this is the first gauge-theoretic proof.