Geologist here! Earth's internal heat budget [1] is a really cool subject since it is surprisingly hard to constrain, yet influences the operation of pretty much the entire solid Earth system (plate tectonics is entirely driven by secular cooling and radiogenic heat production).

Even using the current best estimate that about half of the current 42-46 TW geothermal heat flux is primordial and half is radiogenic, there's a bit of a mismatch if you try to extrapolate back in time* the solution to which may involve either a hotter core than previously expected, anomalously high present heat flux, or more errors in the way we're currently modelling heat loss through the mantle.

*To put it simply, imagine stepping back 1 Myr from the present. The ~half of the heat lost over that period that was _not_ radiogenic came from cooling the core and mantle, so the core and mantle should have been hotter 1 Myr ago. But if the mantle was hotter, it should have been less viscous -> convecting faster -> losing heat even faster in the past. If you keep extrapolating like this, it doesn't take too long for the extrapolated mantle to become physically implausible -- only about 2 or 3 Gyr out of the 4.56 Gyr we have to work with. There are various ways to solve this problem, but we don't know which solution or combination of solutions is what really happened.

Unfortunately there isn't a really good recent review paper on this, but for anyone interested here's a scattering of citations [2-5] that have most of the components to piece together.

[1] https://en.wikipedia.org/wiki/Earth's_internal_heat_budget

[2] doi.org/10.1029/JB085iB05p02517

[3] doi.org/10.1029/2003GL016982

[4] doi.org/10.1016/j.pepi.2020.106457

[5] scholar.google.com/scholar?cluster=7130495919742279897&hl=en