Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2605.07824

Tamed Feynman-Kac diffusion processes: Killing-branching intertwine

Piotr Garbaczewski, Mariusz Żaba

cond-mat.stat-mechmath-phmath.PRnlin.SIquant-phSubmitted May 8, 2026• Announced May 11, 2026• Revised Jun 12, 2026 (v2)
v1 · May 8, 2026v2 · Jun 12, 2026
arXivHTMLPDF

Abstract

Relaxation to equilibrium of a drifted Brownian motion is quantified by a transition probability density function, whose main (multiplicative) entry is an inferred Feynman-Kac kernel of the Schrödinger semigroup operator. Although seemingly devoid of a natural probabilistic significance (except for its explicit path integral definition), the pertinent kernel relaxes to equilibrium as well. The implicit Feynman-Kac potential ${\cal{V}}(x)$, continuous, confining and bounded from below, may take negative values. If positive, ${\cal{V}}(x)$ can be interpreted as the killing rate of the decaying diffusion process. In case of relaxing F-K kernels the killing effects are tamed (often overcompensated). The taming inavoidably appears in conjunction with the existence of the negativity subdomains of ${\cal{V}}(x)$ in $R$. If locally ${\cal{V}}(x) < 0$, its sign inversion $- {\cal{V}}(x)$ can be interpreted as the branching (cloning, alternatively bifurcation) rate in the course of the other wise free random motion. The arising killed diffusion processes with branching, we interpret as the possible path-wise background of tamed (relaxing) Feynman-Kac diffusions. We present acomputer-assisted path-wise arguments, towards a consistency of the killing/branching taming scenario, for a number of nonlinear model systems in one space dimension. Special attention is paid to Feynman-Kac potential shapes in the double well form, where an analytic access to eigenvalues and eigenfunctions is scarce. Throughout the paper the dynamics refers to the positive real time. Since the Newton-type equations of motion for admissible classical trajectories have a Euclidean form (due to the sign inverted force term), we give a brief resume of a couple of their explicit solutions, without recourse to the Euclidean time intuitions, and the instanton lore of related quantum model systems.

Journal reference
(De Gruyter) Open Transport 2026; 1(1): 20260019
DOI
10.1515/ot-2026-0019
Comments
34 pp. 26 figures, invited paper submitted to (De Gruyter Brill, Berlin) "Open Transport"

Related papers

Nonnegative Feynman-Kac Kernels in Schrödinger's Interpolation Problem

Ph. Blanchard, Piotr Garbaczewski, Robert Olkiewicz

The existing formulations of the Schrödinger interpolating dynamics, which is constrained by the prescribed input-output statistics data, utilize strictly positive Feynman-Kac kernels. This implies that the related Markov diffusion processes admit vanishing probability densities only at the boundaries of the spatial volume confining the process. We extend the framework to encompass singular potentials and associated nonnegative Feynman-Kac-type kernels. It allows to deal with general nonnegative solutions of the Schrödinger boundary data problem. The resulting stochastic processes are capable of both developing and destroying nodes (zeros) of probability densities in the course of their evolution.

arXiv:quant-ph/9507002•Jul 5, 1995•2 citations

(Nonequilibrium) Dynamics of Diffusion Processes with Non-conservative Drifts

Piotr Garbaczewski, Mariusz Żaba

Abstract The nonequilibrium Fokker–Planck dynamics with a non-conservative drift field, in dimension $$N\ge 2$$ N ≥ 2 , can be related with the non-Hermitian quantum mechanics in a real scalar potential Vand in a purely imaginary vector potential -iAof real amplitude A (Mazzolo and Monthus in Phys Rev E 107:014101, 2023). Since Fokker–Planck probability density functions may be obtained by means of Feynman’s path integrals, the previous observation points towards a general issue of “magnetically affine” propagators, possibly of quantum origin, in real and Euclidean time. In below we shall follow the $$N=3$$ N = 3 “magnetic thread”, within which one may keep under a computational control formally and conceptually different implementations of magnetism (or surrogate magnetism) in the dynamics of diffusion processes. We shall focus on interrelations (with due precaution to varied, not evidently compatible, notational conventions) of: (i) the pertinent non-conservatively drifted diffusions, (ii) the classic Brownian motion of charged particles in the (electro)magnetic field, (iii) diffusion processes arising within so-called Euclidean quantum mechanics (which from the outset employs non-Hermitian “magnetic” Hamiltonians), (iv) limitations of the usefulness of the Euclidean map $$\exp (-itH_{quant}) \rightarrow \exp (-tH_{Eucl})$$ exp ( - i t H quant ) → exp ( - t H Eucl ) , regarding the probabilistic significance of inferred (path) integral kernels in the description of diffusion processes.

arXiv:2302.10154•Journal of Statistical Physics•May 30, 2024•2 citations

Branching stochastic mechanics: Clustering and connected correlations within a branching-process representation of the Schrödinger equation

Eric Dumonteil, Benoît Bischoff, Alain Letourneau +2 more

Can finite-range correlations hide in the statistics of an extended quantum state? The Schrödinger-Nagasawa transform represents the wave function by positive forward and backward diffusion fields whose product is the Born density. We promote them to branching superprocesses, \(Φ_F\) and \(Φ_B\): diffusion samples stochastic paths, whereas Bohm/Fisher-controlled branching generates genealogies of alternative continuations. With rates evaluated on the prescribed Born density, their means reproduce Schrödinger dynamics exactly. The connected sector exhibits supercritical, critical, and subcritical clustering in confinement and has critical dimension \(d_c=2\) in free space. Stationary eigenmodes remain extended; subcritical clusters acquire the reduced de~Broglie scale. We then let the branching rate respond to the fluctuating product \(Φ_FΦ_B\). In the reciprocal basis, the fields equal a smooth reference \(R\) plus centered fluctuations \(ψ_F,ψ_B\), defining the signed kernel \(C_{\rm FB}(x,y)=\mathbb E_ω[ψ_F(x)ψ_B(y)]\). On the anticorrelated branch, \(ρ_{\rm BSM}(x)=-C_{\rm FB}(x,x)\) is the positive organized connected weight. Stationary matching of this weight to the Born profile is imposed as a closure condition, while off-diagonal decay defines the correlation range. The pair equation splits into collective and relative sectors: spectral cancellation selects the Born collective mode, while suppression of the leading density fluctuation selects the anticorrelated source channel. With relative diffusivity \(D_{\rm eff}\) and positive relaxation rate \(μ_{\rm FB}\), correlations have screening length \(ξ_{\rm FB}=\sqrt{D_{\rm eff}/μ_{\rm FB}}\). Within this reciprocal closure, an extended Born-shaped connected weight and a finite correlation range can coexist in one stochastic kernel.

cond-mat.stat-mechCross-listedarXiv:2608.29807•Sep 1, 2026•1 citation

Many-body diffusion and path integrals for identical particles

L. F. Lemmens, F. Brosens, J. T. Devreese

For distinguishable particles it is well known that Brownian motion and a Feynman-Kac functional can be used to calculate the path integral (for imaginary times) for a general class of scalar potentials. In order to treat identical particles, we exploit the fact that this method separates the problem of the potential, dealt with by the Feynman-Kac functional, from the process which gives sample paths of a noninteracting system. For motion in one dimension, we emphasize that the permutation symmetry of the identical particles completely determines the domain of Brownian motion and the appropriate boundary conditions: absorption for fermions, reflection for bosons. Further analysis of the sample paths for motion in three dimensions allows us to decompose these paths into a superposition of one-dimensional sample paths. This reduction expresses the propagator (and consequently the energy and other thermodynamical quantities) in terms of well-behaved one-dimensional fermion and boson diffusion processes and the Feynman-Kac functional.

arXiv:cond-mat/9601079•Physical review. E, Statistical physics, plasmas, fluids, and related interdisciplinary topics•May 1, 1996•22 citations

Feynman–Kac kernels in Markovian representations of the Schrödinger interpolating dynamics

Piotr Garbaczewski, Robert Olkiewicz

Probabilistic solutions of the so-called Schrödinger boundary data problem provide for a unique Markovian interpolation between any two strictly positive probability densities designed to form the input–output statistics data for the process taking place in a finite-time interval. The key issue is to select the jointly continuous in all variables positive Feynman–Kac kernel, appropriate for the phenomenological (physical) situation. We extend the existing formulations of the problem to cases when the kernel is not a fundamental solution of a parabolic equation, and prove the existence of a continuous Markovian interpolation in this case. Next, we analyze the compatibility of this stochastic evolution with the original parabolic dynamics, which is assumed to be governed by the temporally adjoint pair of (parabolic) partial differential equations, and prove that the pertinent random motion is a diffusion process. In particular, in conjunction with Born’s statistical interpretation postulate in quantum theory, we consider stochastic processes which are compatible with the Schrödinger picture quantum evolution.

arXiv:quant-ph/9505012•Journal of Mathematical Physics•Feb 1, 1996•29 citations

Stochastic motions of the two-dimensional many-body delta-Bose gas, III: Path integrals

Chen, Yu-Ting

This paper is the third in a series devoted to constructing stochastic motions for the two-dimensional $N$-body delta-Bose gas for all integers $N\geq 3$ and establishing the associated Feynman-Kac-type formulas. The main results here prove the Feynman-Kac-type formulas by using the stochastic many-$δ$ motions from [7] as the underlying diffusions. The associated multiplicative functionals show a new form and are derived from the analytic solutions of the two-dimensional $N$-body delta-Bose gas obtained in [4]. For completeness, the main theorem includes the formula for $N=2$, which is a minor modification of the Feynman--Kac-type formula proven in [5] for the relative motions.

arXiv:2505.03006•May 5, 2025•5 citations
Relaxation to equilibrium of a drifted Brownian motion is quantified by a transition probability density function, whose main (multiplicative) entry is an inferred Feynman-Kac kernel of the Schrödinger semigroup operator. Although seemingly devoid of a natural probabilistic significance (except for its explicit path integral definition), the pertinent kernel relaxes to equilibrium as well. The implicit Feynman-Kac potential V(x){\cal{V}}(x)V(x), continuous, confining and bounded from below, may take negative values. If positive, V(x){\cal{V}}(x)V(x) can be interpreted as the killing rate of the decaying diffusion process. In case of relaxing F-K kernels the killing effects are tamed (often overcompensated). The taming inavoidably appears in conjunction with the existence of the negativity subdomains of V(x){\cal{V}}(x)V(x) in RRR. If locally V(x)<0{\cal{V}}(x) < 0V(x)<0, its sign inversion −V(x)- {\cal{V}}(x)−V(x) can be interpreted as the branching (cloning, alternatively bifurcation) rate in the course of the other wise free random motion. The arising killed diffusion processes with branching, we interpret as the possible path-wise background of tamed (relaxing) Feynman-Kac diffusions. We present acomputer-assisted path-wise arguments, towards a consistency of the killing/branching taming scenario, for a number of nonlinear model systems in one space dimension. Special attention is paid to Feynman-Kac potential shapes in the double well form, where an analytic access to eigenvalues and eigenfunctions is scarce. Throughout the paper the dynamics refers to the positive real time. Since the Newton-type equations of motion for admissible classical trajectories have a Euclidean form (due to the sign inverted force term), we give a brief resume of a couple of their explicit solutions, without recourse to the Euclidean time intuitions, and the instanton lore of related quantum model systems.
The existing formulations of the Schrödinger interpolating dynamics, which is constrained by the prescribed input-output statistics data, utilize strictly positive Feynman-Kac kernels. This implies that the related Markov diffusion processes admit vanishing probability densities only at the boundaries of the spatial volume confining the process. We extend the framework to encompass singular potentials and associated nonnegative Feynman-Kac-type kernels. It allows to deal with general nonnegative solutions of the Schrödinger boundary data problem. The resulting stochastic processes are capable of both developing and destroying nodes (zeros) of probability densities in the course of their evolution.
Abstract The nonequilibrium Fokker–Planck dynamics with a non-conservative drift field, in dimension N≥2N\ge 2N≥2 N ≥ 2 , can be related with the non-Hermitian quantum mechanics in a real scalar potential Vand in a purely imaginary vector potential -iAof real amplitude A (Mazzolo and Monthus in Phys Rev E 107:014101, 2023). Since Fokker–Planck probability density functions may be obtained by means of Feynman’s path integrals, the previous observation points towards a general issue of “magnetically affine” propagators, possibly of quantum origin, in real and Euclidean time. In below we shall follow the N=3N=3N=3 N = 3 “magnetic thread”, within which one may keep under a computational control formally and conceptually different implementations of magnetism (or surrogate magnetism) in the dynamics of diffusion processes. We shall focus on interrelations (with due precaution to varied, not evidently compatible, notational conventions) of: (i) the pertinent non-conservatively drifted diffusions, (ii) the classic Brownian motion of charged particles in the (electro)magnetic field, (iii) diffusion processes arising within so-called Euclidean quantum mechanics (which from the outset employs non-Hermitian “magnetic” Hamiltonians), (iv) limitations of the usefulness of the Euclidean map exp⁡(−itHquant)→exp⁡(−tHEucl)\exp (-itH_{quant}) \rightarrow \exp (-tH_{Eucl})exp(−itHquant​)→exp(−tHEucl​) exp ( - i t H quant ) → exp ( - t H Eucl ) , regarding the probabilistic significance of inferred (path) integral kernels in the description of diffusion processes.
Can finite-range correlations hide in the statistics of an extended quantum state? The Schrödinger-Nagasawa transform represents the wave function by positive forward and backward diffusion fields whose product is the Born density. We promote them to branching superprocesses, ΦFΦ_FΦF​ and ΦBΦ_BΦB​: diffusion samples stochastic paths, whereas Bohm/Fisher-controlled branching generates genealogies of alternative continuations. With rates evaluated on the prescribed Born density, their means reproduce Schrödinger dynamics exactly. The connected sector exhibits supercritical, critical, and subcritical clustering in confinement and has critical dimension dc=2d_c=2dc​=2 in free space. Stationary eigenmodes remain extended; subcritical clusters acquire the reduced de~Broglie scale. We then let the branching rate respond to the fluctuating product ΦFΦBΦ_FΦ_BΦF​ΦB​. In the reciprocal basis, the fields equal a smooth reference RRR plus centered fluctuations ψF,ψBψ_F,ψ_BψF​,ψB​, defining the signed kernel CFB(x,y)=Eω[ψF(x)ψB(y)]C_{\rm FB}(x,y)=\mathbb E_ω[ψ_F(x)ψ_B(y)]CFB​(x,y)=Eω​[ψF​(x)ψB​(y)]. On the anticorrelated branch, ρBSM(x)=−CFB(x,x)ρ_{\rm BSM}(x)=-C_{\rm FB}(x,x)ρBSM​(x)=−CFB​(x,x) is the positive organized connected weight. Stationary matching of this weight to the Born profile is imposed as a closure condition, while off-diagonal decay defines the correlation range. The pair equation splits into collective and relative sectors: spectral cancellation selects the Born collective mode, while suppression of the leading density fluctuation selects the anticorrelated source channel. With relative diffusivity DeffD_{\rm eff}Deff​ and positive relaxation rate μFBμ_{\rm FB}μFB​, correlations have screening length ξFB=Deff/μFBξ_{\rm FB}=\sqrt{D_{\rm eff}/μ_{\rm FB}}ξFB​=Deff​/μFB​​. Within this reciprocal closure, an extended Born-shaped connected weight and a finite correlation range can coexist in one stochastic kernel.
For distinguishable particles it is well known that Brownian motion and a Feynman-Kac functional can be used to calculate the path integral (for imaginary times) for a general class of scalar potentials. In order to treat identical particles, we exploit the fact that this method separates the problem of the potential, dealt with by the Feynman-Kac functional, from the process which gives sample paths of a noninteracting system. For motion in one dimension, we emphasize that the permutation symmetry of the identical particles completely determines the domain of Brownian motion and the appropriate boundary conditions: absorption for fermions, reflection for bosons. Further analysis of the sample paths for motion in three dimensions allows us to decompose these paths into a superposition of one-dimensional sample paths. This reduction expresses the propagator (and consequently the energy and other thermodynamical quantities) in terms of well-behaved one-dimensional fermion and boson diffusion processes and the Feynman-Kac functional.
Probabilistic solutions of the so-called Schrödinger boundary data problem provide for a unique Markovian interpolation between any two strictly positive probability densities designed to form the input–output statistics data for the process taking place in a finite-time interval. The key issue is to select the jointly continuous in all variables positive Feynman–Kac kernel, appropriate for the phenomenological (physical) situation. We extend the existing formulations of the problem to cases when the kernel is not a fundamental solution of a parabolic equation, and prove the existence of a continuous Markovian interpolation in this case. Next, we analyze the compatibility of this stochastic evolution with the original parabolic dynamics, which is assumed to be governed by the temporally adjoint pair of (parabolic) partial differential equations, and prove that the pertinent random motion is a diffusion process. In particular, in conjunction with Born’s statistical interpretation postulate in quantum theory, we consider stochastic processes which are compatible with the Schrödinger picture quantum evolution.
This paper is the third in a series devoted to constructing stochastic motions for the two-dimensional NNN-body delta-Bose gas for all integers N≥3N\geq 3N≥3 and establishing the associated Feynman-Kac-type formulas. The main results here prove the Feynman-Kac-type formulas by using the stochastic many-δδδ motions from [7] as the underlying diffusions. The associated multiplicative functionals show a new form and are derived from the analytic solutions of the two-dimensional NNN-body delta-Bose gas obtained in [4]. For completeness, the main theorem includes the formula for N=2N=2N=2, which is a minor modification of the Feynman--Kac-type formula proven in [5] for the relative motions.