A question about generalized nonvanishing
arXiv:2312.09441
Abstract
Let be a projective, -factorial log canonical pair and let be a pseudoeffective -divisor on such that is pseudoeffective. Is there an effective -divisor on such that is numerically equivalent to ? We are not aware of any counterexamples, but the answer is not completely clear even in the case of surfaces.
Final version, to appear in Rendiconti del Circolo Matematico di Palermo