The paper studies a partial differential equation that models how a quantity (think of it as a temperature or chemical concentration) evolves over time and space. The equation has two competing forces: a reaction term that drives the solution to grow explosively (blow-up), and an absorption term that depends on how steeply the solution varies in space (the gradient). The central question is whether the absorbing term is strong enough to always prevent the solution from becoming infinite in finite time, provided the initial state dies off far away from the origin.
This had been an open problem for many years. The authors resolve it by proving that, yes, blow-up is completely prevented when the initial data decays to zero at spatial infinity and the absorption exponent is at least as large as the reaction exponent. The key technical tool they develop is a new type of gradient estimate, called a Bernstein-type estimate, which gives a universal bound on how steep the solution can become. This estimate is novel because it only works in the parameter range where absorption is strong relative to reaction, and it holds independently of the specific solution or initial data, making it potentially useful in other problems too.
The authors also show that the decay condition on the initial data is essentially the sharpest possible. If instead the initial data stays positive (or approaches a positive value) far from the origin, blow-up does happen in finite time. They even characterize how the solution behaves near the blow-up moment when this occurs. Together, these results draw a clean boundary between safe and dangerous initial conditions, fully settling the long-standing question about this class of equations.