Wetterich’s Equation and its Boundary Conditions for Radon Measures on Locally Convex Spaces
Jobst Ziebell 0000-0002-9715-6356 Faculty of Mathematics and Computer Science, Friedrich-Schiller-University, Jena, Germany · 2025
Read Wetterich’s Equation and its Boundary Conditions for Radon Measures on Locally Convex Spaces by Jobst Ziebell 0000-0002-9715-6356 Faculty of Mathematics and Computer Science, Friedrich-Schiller-University, Jena, Germany (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-01742.txt.
Introduction
Jobst Ziebell studies Wetterich's equation for Radon measures on locally convex spaces. The equation drives flows of effective average actions in field theory. It is used widely to probe quantum field traits. Its domain and boundary terms need care in infinite dimensions.
Ziebell shows the equation holds for Radon measures under set conditions. The effective action's domain is the Lusin affine kernel of the measure. Firm ground in this generality supports functional renormalization work on complex field theories where standard spaces fail. Locally convex spaces cover most function spaces physics uses, so the result's scope reaches far past the toy cases that motivated it.
Catalog record
| CREATOR | Jobst Ziebell 0000-0002-9715-6356 Faculty of Mathematics and Computer Science, Friedrich-Schiller-University, Jena, Germany |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 101 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-01742
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Wetterich’s Equation and its Boundary Conditions for Radon Measures on Locally Convex Spaces Jobst Ziebell 0000-0002-9715-6356 Faculty of Mathematics and Computer Science, Friedrich-Schiller-University, Jena, Germany Abstract Wetterich’s equation and corresponding flows of effective average actions are used frequently in theoretical physics to study the properties of quantum field theories. Under appropriate conditions, Wetterich’s equation also holds for Radon measures on locally convex spaces and the domain of the effective average action is the Lusin affine kernel of the measure. The resulting flow interpolates between the convex conjugate of the cumulant-generating function of the measure in question and its (generalised) Onsager-Machlup function. The underlying metric of the latter is induced by a family of measurable bilinear functionals that can be understood as bilinear versions of Lusin measurable linear functionals. \recalctypearea 1 Introduction In theoretical physics, one studies the flow of ‘effective average actions’ which in the case of a real-valued scalar field are real-valued functions \Gamma_{k} on an affine space [ 10 ] . The asymptotic boundary condition for large flow parameters should correspond to the ‘classical action’ . In [ 23 ] , the author showed that this picture indeed holds for certain regularised models, encoded as measures that are (measure-theoretically) equivalent to certain Gaußian measures. Then the asymptotic boundary condition becomes the sum of the squared Cameron-Martin norm and the logarithm of the given density. It is an intriguing question, whether the flow equation also holds for non-regularised quantum field theories and how the boundary conditions behave. From a purely mathematical point of view, one can think of taking a Radon measure on a locally convex space and use Wetterich’s equation as a tool to calculate its Onsager-Machlup function. The natural domain of the effective average action then becomes the Lusin affine kernel of the measure. While not touched upon in this paper, the reverse programme is also of considerable interest: Given some function on an affine space, can it be the Onsager-Machlup function of a measure? In theretical physics, this is exactly the problem of quantising a given (Euclidean) classical action functional. It is well-established that this is possible in e.g. two spacetime dimensions for a wide class of actions [ 19 ] and a landmark result also produced the celebrated \Phi^{4}_{3} model (see e.g. [ 12 ] for a short summary and a list of original references). At the same time, it is known - though from a completely different quantisation ansatz - that many action functionals can only correspond to ‘trivial’ quantum field theories, that is to Gaußian measures [ 12 , Section 21.6] . Ideally, the existence of a solution to the flow equation with given boundary conditions should be able to answer the question of the triviality of a given action functional. A final motivation of this topic, also not touched upon in this paper, is the necessity of ‘renormalisation’ , which in this setting can be summarised as follows: Suppose a sequence (\mu_{n})_{n\in\mathbb{N}} of Radon probability measures are given that satisfy the flow equation with boundary conditions given by a suitable sequence (f_{n})_{n\in\mathbb{N}} of functions. If \mu_{n} converges weakly (or perhaps in some other sense) to a Radon measure \mu , in what sense do the boundary conditions f_{n} converge? From a physical point of view, \lim_{n\to\infty}f_{n} is typically ill-defined, thanks to the divergence of certain parameters (coupling constants) in f_{n} . However, the hope of many theoretical physicist - including the author - is that the flow (\Gamma_{k})_{k\geq 0} of the effective average action of the limit measure \mu captures these divergences by making the boundary condition \lim_{k\to\infty}\Gamma_{k}\left(\frac{y}{k}\right) (1) well-defined (the factor 1/k comes from power counting but could in general be something else). That is, even if the boundary condition \lim_{n\to\infty}f_{n} is ill-defined, there should be a way to capture the asymptotics of it. In fact, one may hope that for the corresponding flows (\Gamma^{n}_{k})_{k\geq 0} of the measures \mu_{n} respectively, at finite k\geq 0 \lim_{n\to\infty}\Gamma^{n}_{k}=\Gamma_{k} (2) in a suitable sense. 2 Preliminaries \overline{\mathbb{R}}=\mathbb{R}\cup\{-\infty\}\cup\{\infty\} and for a function f:I\to\mathbb{R} on an interval I\subseteq\mathbb{R} , denote by D^{+}f(x)=\limsup_{y\searrow x}[f(y)-f(x)]/(y-x)\in\overline{\mathbb{R}} the upper right Dini derivative at a point x\in I with x<\sup I . Similarly, D_{-}f(x) denotes the lower left Dini derivative at x . Lemma 2.1 . Let a<b and f:[a,b]\to\mathbb{R} be a continuous, monotonically increasing function with D^{+}f(x)\in\mathbb{R} for every x\in[a,b) . Then f is absolutely continuous. Proof. Let \lambda denote the Lebesgue measure on [a,b] . From [ 17 , p. 272] , D^{+}f is \lambda -measurable and we have \lambda\left(f\left(B\right)\right)\leq\int_{B}\left|D^{+}f\right|\mathrm{d}\lambda\,, (3) for every \lambda -measurable set B\subseteq[a,b] . In particular, it follows that f takes \lambda -null sets to \lambda -null sets such that the claim follows from [ 16 , Theorem 7.18] . ∎ All vector spaces in this paper are considered over the field of real numbers and the topologies considered are not assumed to be Hausdorff unless explicitly stated. The complexification of a vector space X is denoted by X_{\mathbb{C}} . If X is equipped with some topology \tau (not necessarily a vector topology), we define (X,\tau)^{*} to be the set of all linear real-valued functionals on X that are upper semicontinuous with respect to \tau . It is easy to see that (X,\tau)^{*} is a cone , i.e. it is closed under linear combinations with nonnegative coefficients. Note that in the case of a vector topology, (X,\tau)^{*} is precisely the usual topological dual and a vector space in its own right. We also define the conjugate topology \bar{\tau}=\left\{-U:U\in\tau\right\} (4) and it is straightforward to see that (X,\tau)^{*}=-(X,\bar{\tau})^{*} . We shall consider asymmetrically normed spaces which are pairs (X,p) where X is a vector space and p is an asymmetric norm on X , i.e. a function p:X\to[0,\infty) such that p\left(x\right)=p\left(-x\right)=0\iff x=0\,,\qquad p\left(rx\right)=rp\left(x\right)\,,\qquad p\left(x+y\right)\leq p\left(x\right)+p\left(y\right) (5) for all x,y\in X and r\geq 0 . The asymmetric norm induces a first-countable (not necessarily Hausdorff) topology for which sets of the form x+p^{-1}([0,\epsilon)) for x\in X and \epsilon>0 form a neighbourhood basis at any given x\in X . To such a pair (X,p) , we associate the conjugate asymmetrically normed space (X,\bar{p}) with \bar{p}(x)=p(-x) and remark that the induced topology is indeed the corresponding conjugate topology. We also equip (X,p)^{*} with the functional p^{*}:(X,p)^{*}\to[0,\infty) given by \phi\mapsto\sup\{\phi(x)|x\in X:p(x)\leq 1\} . p^{*} is a cone norm in the sense that p^{*}\left(\phi\right)=0\iff\phi=0\,,\qquad p^{*}\left(r\phi\right)=rp^{*}\left(\phi\right)\,,\qquad p^{*}\left(\phi+\psi\right)\leq p^{*}\left(\phi\right)+p^{*}\left(\psi\right) (6) for all \phi,\psi\in(X,p)^{*} and r\geq 0 . The pair ((X,p)^{*},p^{*}) is then a normed cone and every x\in X induces a continuous functional on it with \phi\mapsto x(\phi) that is linear in the sense of the usual compatibility with respect to linear combinations with non-negative scalars [ 7 , Proposition 2.4.22] . The above notation is consistent with the conventional normed setting, i.e. (X,p)^{*} is just the ordinary continuous dual space whenever p is a norm on X and then p^{*} is the conventional dual norm. In either case p^{*} similarly induces a first-countable topology on (X,p)^{*} . Furthermore, we consider a weak topology on (X,p) , where a net (x_{\alpha})_{\alpha\in I} in X converges p -weakly to x\in X , if \limsup_{\alpha}\phi(x_{\alpha})\leq\phi(x) for all \phi\in(X,p)^{*} . Again, if p is a norm, the weak topology is the usual weak vector topology. Lemma 2.2