In the literature, two different definitions of the free energy are common:
- thermodynamic / probabilistic free energy
- geometric free energy
The question is how to interprete the delta function in the following definition [1]
F(s) = - \beta^{-1} \ln \int_{R^n} \exp(-\beta V(x)) \delta(s)
What are their differences/similarities?