← All problems

The rational-slice versus biprojectively H-slice question

For every knot K⊂S3K\subset S^3, if KK is rationally slice— that is, KK bounds a smoothly properly embedded disk in a rational homology 44-ball—then KK is biprojectively H-slice.

References

Progress summary

Refreshed
Open · solution reported

A 2026 preprint claims to separate rational sliceness from the stronger BPH-sliceness condition by producing infinitely many counterexamples.

The question asks whether every rationally slice knot is biprojectively H-slice, answering a question attributed to Lee.

Known results

Earlier work of Manolescu and Piccirillo, together with the reported new result, gives a classification of BPH-sliceness for prime knots with at most nine crossings.

2026 claimed resolution

Elise Brod's preprint claims that, for even q≥4q \ge 4 with 3∤q3 \nmid q, the Turk's head knots Th⁡(3,q)\operatorname{Th}(3,q) are rationally slice but not H-slice in either orientation of connected sums of projective planes, hence not BPH-slice. The first example is Th⁡(3,4)=818\operatorname{Th}(3,4)=8_{18}. The argument uses a lattice-embedding obstruction excluding every nn. The broader smooth-sliceness questions remain open; the claim has not received independent mathematical assessment in the retrieved sources.

Current status (as of November 2026): The preprint claims the rational-slice versus BPH-slice question is settled negatively, including 8188_{18}, but this resolution remains unverified and broader sliceness questions remain open.

Sources

Solutions 0

No solutions have been posted yet.