In General Relativity, spacetime itself can be curved by its very nature. Vacuum solutions which are exact solutions of the Einstein Field Equations (the core

In General Relativity, spacetime itself can be curved by its very nature. Vacuum solutions which are exact solutions of the Einstein Field Equations (the core set of partial differential equations of General Relativity; the EFEs), have been the subject of extensive research since the 1920s. A test particle is one with negligible mass and extent, and thus does not itself perturb the curvature of the background vacuum spacetime. If a real particle is electrically neutral, extremely low mass, and nearly pointlike, then in a vacuum spacetime it should behave very much like a test particle. Moreover, for microscopic real objects, which aren't truly pointlike and which may not be wholly chargeless, it is impractically hard to tell at a large enough spatial distance that it's not a test particle. So a test particle or test field does not source the spacetime curvature -- it just rides along the background spacetime, which is arbitrarily curved by its nature. We can extend the notion to a test field that is almost wholly massless, like a single-electron quantum field. The test field sits on the background, but doesn't influence the background's dynamics. Real quantum fields, like test fields, permeate the whole of spacetime. However they tend to have abundant stress-energy in them, so source some of the curvature of the whole spacetime, and typically not uniformly. One can fix a gauge (like how one calibrates an air-pressure altimeter) that lets one talk about the inherent curvature of the background spacetime and the contribution to the total curvature of the lowest density of the quantum field. Typically the background is chosen to be some exact vacuum solution of the EFEs with corrections added via something like perturbation theory; flat spacetime, Schwarzschild and Robertson-Walker expanding spacetimes are popular backgrounds, depending on what one is studying. Sometimes, though one wants to think about an electron rather than some characteristic set of numbers in some all-over-spacetime fields. In that case one (probably implicitly) fixes a gauge that lets one talk about the electron as an independent object that generates an electromagnetic field which, because that has some nonzero stress-energy (zero being set by our gauging choices and being found sufficiently far from our electron) it influences the background curvature of the spacetime (the background again being set by our gauging choices). In either paradigm, one might want to think about, in reverse order, the gravitational back-reaction of the electron (and its EM field)'s gravitational footprint on itself, or the gravitational back-reaction of the overdensities in the all-over-spacetime (classical or quantum) fields on the regions of spacetime they are found in. "Gravity gravitates" is the problem: this arises from the EFEs' non-linearity. Fortunately, for a single electron these back-reactions will be very small, because electrons in either paradigm have very little stress-energy, so their contribution to any curvature will thus be so extremely tiny that it will be practically immeasurable, so there is basically no hope of measuring higher-order contributions (that is, the electron's gravitational influence's own influence on the curvature). Negiligibility saves a lot of headaches, because working with gravitational backreaction is hard.