All-level flowed Drinfel'd–Sokolov reduction for discrete and relaxed modules

Prove the all-level identification of the flowed Drinfel'd–Sokolov cohomology of the affine discrete-series and relaxed principal-series modules with the corresponding Virasoro Verma modules, including the additional level-one quotient required at the $j=k/2$ endpoint to obtain the Virasoro vacuum module.

Background

The appendix establishes the level-zero principal-series cohomology explicitly and verifies the expected cohomology through levels N=0,1,2,3,4N=0,1,2,3,4. The expected pattern is ghost-number-zero cohomology with dimensions equal to the partition numbers, corresponding to the Virasoro descendant tower.

Nevertheless, the authors state that the all-level proof remains unavailable. This unresolved representation-theoretic problem is central to making the quantize-then-reduce construction mathematically complete for both sub-threshold and super-threshold sectors.

References

What fails to adapt straightforwardly from the standard theorems on (unflowed) discrete series are the all-level proofs: but they fail irrespective of level, so the fact that we can show that it works by direct cohomology calculations at low levels can be viewed as indicating that it is the proof methods that are failing, and not the reduction itself. We hope to present a complete demonstration in .

— Building a Quantum Black Hole Microstate: A Bulk Path Integral for a Heavy Virasoro Primary  (2608.14541 - Krishnan et al., 14 Aug 2026) in Appendix B, “Principal-Series Zero Mode and BRST Cohomology”