Pith. sign in

REVIEW 5 major objections 4 minor 21 references

Strong solutions of SDE's with rough coefficients

T0 review · 5 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A merely measurable Markovian drift forces a strong solution of $dX_t=b(t,X_t)dt+dW_t$ whenever the Girsanov exponential has a finite $L^{1+\epsilon}$ moment, and the solution map is H-C regular.

desk verdict Novel upper-floor construction, but the main variational identity is false; the paper as written does not prove the claimed strong existence. read the letter →

arxiv 2507.21592 v2 pith:GBQ6O5MY submitted 2025-07-29 math.PR

Ali Suleyman Ustunel This is my paper · ORCID
classification math.PR MSC 60H0760H1060H3037A3557C7094A17
keywords entropyGirsanovtheoremstrongsolutionsstochasticdifferentialequationsalmostsureinvertibilitycylindricalBrownianmotionabstractWienerspacemeasurabledrift
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proves that the stochastic differential equation $dX_t=b(t,X_t)dt+dW_t$ has a strong, pathwise unique solution when the drift $b$ is only assumed measurable and Markovian, provided the Girsanov density $\rho(-\tilde\delta b)=e^{-f}$ satisfies $\mathbb{E}[\rho(-\tilde\delta b)]=1$ and $\mathbb{E}[\rho(-\tilde\delta b)^{1+c_b}]<\infty$ for some fixed $c_b>0$. No continuity, boundedness, or spatial integrability of $b$ is needed; the entire hypothesis is carried by the exponential martingale that $b$ generates. The proof moves the problem to an abstract Wiener space over $\Omega=C([0,1],W)$ with Cameron-Martin space $H_2=H\otimes_2 H$, where an adapted perturbation of identity $V^{\tau}$ is shown to be almost surely invertible, and its inverse solves the SDE. A variational identification then presents the solution as the unique minimizer of an entropic cost functional, and an approximation argument proves the solution is adapted to the driving Brownian motion, hence strong.

What carries the argument

The central object is the adapted perturbation of identity $V^{\tau}=I_{H_2}+v^{\tau}$ on the upper abstract Wiener space. Its derivative in Cameron-Martin directions is controlled by the bounds $\sup_{t\le\tau}\sup_{\|K\|_2\le M}|\dot v_t(B_t(\omega)+K_t)|_H\le c_{1,M}(\tau,\omega)$ and the analogous bound for $\nabla\dot v_t$, which together give the H-C$^1$ (Hilbert-Cameron $C^1$) property. Because the perturbation is adapted, the modified Carleman-Fredholm determinant of $I_{H_2}+\nabla v^{\tau}$ equals one, so the abstract Wiener-space change-of-variables formula reduces to $\mathbb{E}_P[G\circ V^{\tau}]=\mathbb{E}_P[G N(\cdot,V^{\tau})]$, and the equality of expectations for all $G\in C_b(\Omega)$ forces $N(\cdot,V^{\tau})=1$ almost surely. The H-analyticity of $w\mapsto Q_{1-t}e^{-f}(w)$, obtained through the shifted heat-kernel identity, supplies the regularity that makes the whole construction run, and the variational functional $K(\xi)=\mathbb{E}_P\int_0^1|\dot\xi_{1,s}+b(s,\xi_{1,s}+B_{1,s})|^2ds$ links the inverse of $V^{\tau}$ back to the original SDE.

What would settle it

Check whether estimates (2.4)-(2.5) hold for a drift $b$ satisfying (1.1)-(1.2): if for some such $b$ the gradient $\dot v_t(B_t)=-\nabla\log Q_{1-t}e^{-f}(B_t)$ is unbounded on Cameron-Martin balls of finite radius on a set of positive $P$-measure, then $V^{\tau}$ is not H-C$^1$ and the claimed invertibility is unsupported. A direct computation of the multiplicity $N(\omega,V^{\tau})$ for an admissible rough drift that exceeds one on a positive-measure set would likewise contradict (2.6) and refute Theorem 1.

Watch

Extended reading notes

Core claim

The central claim is that under assumptions (1.1)-(1.2) the SDE $dX_t=b(t,X_t)dt+dW_t$ has a strong solution whose solution map is H-C regular, i.e., Hilbert-Cameron regular in the abstract Wiener space sense, on $( Omega,H_2,P)$ with $\Omega=C([0,1],W)$ and $H_2=H\otimes_2 H$. On this upper floor the adapted perturbation of identity $V^{\tau}_t(B)=B_t(B)+\int_0^{\tau\wedge t}\dot v_s(B_s(B))ds$, with $\dot v_s=-\nabla\log Q_{1-s}e^{-f}$, is proved $P$-a.s. invertible: the H-C$^1$ estimates (2.4)-(2.5) give $v^{\tau}$ the needed regularity, the adaptedness makes the Carleman-Fredholm determinant equal to one, and the change-of-variables formula (2.6) forces the multiplicity $N(\cdot,V^{\tau})$ to equal one almost surely. The inverse $U$ satisfies $dU_t=-\dot v(t,U_t)dt+dB_t$, and the variational argument identifies $X_t=\xi_{1,t}+B_{1,t}$, where $\xi$ is the unique minimizer of the entropic functional, as the solution of the original SDE. Theorem 3 completes the proof by showing $X$ is adapted to the filtration of $B_1$, so the weak solution constructed at the upper floor is in fact strong; the same construction yields real H-analyticity of the solution for $t<1$.

Load-bearing premise

The load-bearing premise is that the shift map $V^{\tau}$ built from the drift is smooth enough in the Cameron-Martin sense that the abstract change-of-variables formula applies and forces its multiplicity to be one almost surely, a regularity step that may fail for Girsanov densities only known to lie in $L^{1+\varepsilon}$.

Editorial extensions

If this is right

  • Any measurable Markovian drift whose Girsanov density has mean one and a fixed $L^{1+c_b}$ moment produces a pathwise unique strong solution of the SDE, without any continuity or growth condition on the drift itself.
  • The solution map is H-C regular on the abstract Wiener space, so the solution varies smoothly with Cameron-Martin shifts rather than only in $L^p$.
  • The variational characterization identifies the strong solution as the unique minimizer of an explicit entropic cost, giving a variational calculus for singular drifts.
  • Strong existence is obtained through almost-sure invertibility of an adapted perturbation of identity, so uniqueness and regularity come from the geometry of the Gaussian space rather than from pathwise estimates on $b$.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Extension the paper leaves implicit: the same upper-floor inversion scheme might prove strong existence for SDEs driven by other Gaussian processes, wherever the heat semigroup $Q_t$ and the H-analyticity of $Q_{1-t}e^{-f}$ survive.
  • A natural test is the critical case $c_b=0$: the proof needs the extra $L^{1+c_b}$ room to obtain the bounds (2.4)-(2.5), so it remains open whether plain $L^1$ integrability of the Girsanov density already forces a strong solution.
  • Because the argument is variational, one could attempt to compute the minimizer explicitly for structured drifts, turning the existence proof into a numerical or analytical recipe for the solution map.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 4 minor

Summary. The paper claims a proof of strong existence and uniqueness for the SDE dX_t = b(t,X_t)dt + dW_t on [0,1] for a measurable, Markovian drift b, under only the assumptions (1.1) that the Girsanov density ε(-δb̃) has expectation 1 and (1.2) that it lies in L^{1+c_b}. The method is to pass to an 'upper floor' abstract Wiener space (Ω,H_2,P), define an adapted perturbation of identity V^τ from the logarithmic gradient of the heat semigroup, assert its almost-sure invertibility via a change-of-variables formula with multiplicity one, identify its inverse as the strong solution of a related SDE, and finally use a variational calculation to show that the minimizer solves the original SDE and is adapted to the driving Brownian motion B_1.

Significance. If the result were correct, it would be a striking extension of strong-existence theory for SDEs with rough drifts, reducing the hypothesis to an integrability condition on the Girsanov density and introducing a genuinely new abstract-Wiener-space technique. The paper builds on the author's substantial prior work on invertibility of adapted shifts and variational calculus, and the overall architecture is ambitious and worth serious study. However, the proof as written contains load-bearing algebraic and analytic gaps that prevent the central claims from being established; in particular, the variational identification in §3 is algebraically incorrect, and the invertibility step in §2 is asserted rather than proved under the stated hypotheses.

major comments (5)
  1. [§3, after (3.8)] The variational computation contains a false algebraic identity. With N=||ξ||²_H2, M=∫|ξ̇_{1,s}|²ds, D=∫b·ξ̇_{1,s}ds and C=∫|b|²ds, the expression 1/2(N+∫|ξ̇_{1,s}+b|²ds)-1/2M equals 1/2N+D+1/2C, while the preceding line for K(ξ) is 1/2N+D+C. The displayed chain therefore already loses a factor 1/2C. Even if this were repaired, the final equality with EP∫|ξ̇_{1,s}+b|²ds would require N=M, which is false in general: for b=0 and the deterministic two-parameter function k(t,s)=(t-1)²ts, one has ξ̇_{1,s}=0 and hence the paper's final expression is 0, whereas K=1/2||k||²_H2>0. Consequently the extraction of the SDE dX_t=-b(t,X_t)dt+dB_{1,t} from the minimizer is unsupported.
  2. [§2, proof of Theorem 1] The almost-sure invertibility of V^τ is asserted from the change-of-variables formula (2.6) without verifying the hypotheses of that formula. The estimates (2.4)-(2.5) are claimed to follow from H-analyticity and the strict positivity of the martingale, but no argument is given that these bounds are uniform over the H-balls needed for the H-C^1 property of v^τ on (Ω,H_2,P). Moreover, the conclusion N(·,V^τ)=1 P-a.s. from EP[G∘V^τ Λ]=EP[G] presupposes the area formula and that the multiplicity function is well-defined; this is exactly the kind of regularity that must be proved before use.
  3. [§2, martingale representation] The representation of the martingale EP[e^{-f}∘B_1|B_t] as the exponential martingale exp(-∫(v̇_s,dB_s)-1/2∫|v̇_s|²_H ds) requires an Itô formula for the W-valued cylindrical Brownian motion and sufficient regularity of the heat semigroup map w↦Q_{1-t}(e^{-f})(w). Under the sole assumption e^{-f}∈L^{1+ε}(μ), the H-differentiability and the integrability of the stochastic integral are not justified; the paper does not state or prove the precise conditions under which this representation holds.
  4. [§3, stochastic integral term] The vanishing of the term EP[∫ b(s, ξ_{1,s}+B_1(s)) dB_1(s)] is asserted from orthogonality of Brownian increments, but the integrand is not shown to be adapted to the filtration of B_1. The adaptedness of the minimizer is one of the main conclusions of the paper and cannot be assumed in the variational computation. The argument therefore contains a circularity: it uses the B_1-adaptation of ξ to cancel the stochastic integral before proving that adaptation.
  5. [Theorem 3] The approximation argument via e^{-f_n}=P_{1/n}E[e^{-f}|V_n] and the convergence v̇_n→v̇ P-a.s. are asserted without proof. The weak-convergence passage showing that U^1 is adapted to the filtration of B_1 relies on a limiting argument for which the necessary tightness and continuity in the SDE coefficients are not demonstrated; convergence of the pair (U_n,B_1) in the weak sense does not by itself preserve B_1-adaptedness of the limit. Thus the strong-solution claim of Theorem 3 is not rigorously established.
minor comments (4)
  1. [Abstract] The phrase 'without no regularity hypothesis' should read 'without any regularity hypothesis'.
  2. [p. 3, proof of Theorem 1] The term 'Carlman-Fredholm determinant' is a typo; it should be 'Carleman-Fredholm determinant'.
  3. [p. 4, reference [5]] The conference proceedings 'Proc. 4 th Berkley Sym. Math. Stat. Prob.' misspells 'Berkeley'.
  4. [p. 4, chain in Theorem 2 proof] The chain EP[sup_{t<1}||U_t||_W] = ... = EP[sup_{t≤1}||B_t||_W e^{-f∘B_1}] contains the typo 'ρ(-δB(v)' and the notation for the change of measure is inconsistent.

Circularity Check

2 steps flagged · score 6.0 of 10

The §3 variational chain reduces the minimizer's SDE to an asserted norm identity — ‖ξ‖²_H2 = ∫|ξ̇_{1,s}|²ds — that contradicts the paper's own H2 definition and fails already for b=0, while the underlying 'minimum ⇔ strong solution' equivalence is imported from the author's prior [19,20] with only an 'extend easily' footnote.

  1. uniqueness imported from authors [Section 3, eq. (3.8) and footnote 1]
    "We know that the strong existence of the SDE (2.7) is equivalent to the existence of the unique minimizing element of the following variational problem (cf., [19, 20]) 1: (3.8) inf( K(ξ) : ξ ∈ L2 a(P, H2)) = inf( EP [1/2 ∥ξ∥2 H2 + f ◦ B1 ◦ (IB + ξ)] : ξ ∈ L2 a(P, H2)) = 0. Footnote: In fact in these references the classical Wiener space has been treated but the results extend easily to our abstract Wiener space (Ω, H2, P) with the same proof due to stochastic integral representation of the (Bt, t∈ [0, 1])-martingales."

    The premise that strong existence of (2.7) is equivalent to existence of a unique minimizer of (3.8) with infimum 0 is the load-bearing bridge of §3 and Theorem 3: it is exactly what lets the author read the strong SDE off the minimizer. The equivalence is not proved in this paper; it is cited to the author's own earlier works [19,20] on the classical Wiener space, and the transfer to the new upper-floor abstract Wiener space (Ω,H2,P) is dismissed in a footnote as 'extend easily ... with the same proof.' The target claim is thus derived from a self-cited equivalence whose new-case version is asserted, not established: the minimizer is declared the forced strong solution by the author's prior theorem rather than by an argument given here.

  2. other [Section 3, displayed computation after (3.8) and the following paragraph]
    "= 1/2 EP [∥ξ∥2 H2 + ∫_0^1 |ξ̇1,s + b(s, ξ1,s(B) + B1(s))|2 ds] − 1/2 EP [∫_0^1 |ξ̇1,s|2 ds] = EP [∫_0^1 |ξ̇1,s + b(s, ξ1,s(B) + B1(s))|2 ds]. If ξ is the unique minimizer of K, since E[e−f ] = 1, letting, Xt = ξ1,t + B1(t) we realize that X satisfies the SDE dXt = −b(t, Xt)dt + dB1,t almost surely."

    The final equality reduces K(ξ) to EP∫|ξ̇1,s + b|2 ds by completing the square, which requires ‖ξ‖2 H2 = ∫_0^1 |ξ̇1,s|2H ds. But §3 defines ‖ξ‖2 H2 = EP∫_0^1 |ξ̇(t)|2H dt, an integral over all upper times t of an H-norm that is itself an integral over lower times s, not the terminal slice t=1. The reduction is false already for b=0: for the deterministic adapted element k(t,s)=(t−1)2t2s, ∂t k(1,·)=0, so the claimed final value is 0, while K(k)=1/2‖k‖2H2=1/1260>0. The sentence 'letting Xt = ξ1,t + B1(t) we realize that X satisfies the SDE' extracts the strong solution from this identity; the SDE conclusion is forced through an asserted reduction that contradicts the paper's own definition of the H2 norm.

full rationale

Two load-bearing reductions are asserted rather than proved. (1) Section 3 opens with the equivalence 'strong existence of (2.7) ⇔ unique minimizer of (3.8) with infimum 0', cited to the author's own [19,20], and the footnote concedes that those references treat only the classical Wiener space, declaring that the results 'extend easily' to the upper-floor abstract Wiener space (Ω,H2,P) 'with the same proof'; the whole variational identification, and hence Theorem 3, rests on this self-imported equivalence with the new-case transfer unproved. (2) The displayed chain after (3.8) collapses 1/2‖ξ‖2H2 + 1/2∫|ξ̇1,s + b|2 ds − 1/2∫|ξ̇1,s|2 ds to ∫|ξ̇1,s + b|2 ds, which requires ‖ξ‖2H2 = ∫|ξ̇1,s|2 ds; under the definition of ‖·‖H2 given in the same section this is not an identity and fails for b=0 with k(t,s)=(t−1)2t2s, so the 'realization' that the minimizer solves dX = −b dt + dB1 is read off a false reduction. Separately, Theorem 1's multiplicity claim that N(·,Vτ) 'happens to be equal to one' from (2.6) is an unsupported inference (E[N]=1 does not force N=1 without surjectivity of Vτ), a correctness gap rather than a circularity; and the change-of-variables formula cited to [21] is an externally established general result, so that citation is not itself circular. Because the central strong-existence conclusion is extracted from an identity that contradicts the paper's own definitions and from an equivalence imported without proof from the author's prior work, the derivation chain is partially circular.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper's central claim rests on standard Wiener space background, on previous results of the same author for variational calculus and invertibility of adapted shifts, and on a set of asserted analytic bounds for the logarithmic gradient of the Mehler semigroup. No free parameters are fit to data.

assumptions (4)
  • domain assumption Girsanov integrability hypotheses (1.1) and (1.2) are the only regularity imposed on b.
    The proof never uses differentiability of b; it only uses finiteness of the Girsanov exponential moments.
  • standard math Change-of-variables formula on abstract Wiener space for H-C^1 adapted perturbations with multiplicity (ref [21], Ch. 3).
    Invoked in (2.6) to show measure preservation and multiplicity one for V^τ; the paper cites [21] instead of proving it.
  • domain assumption Equivalence of strong existence for (2.7) with existence of a unique minimizer of the variational functional K (refs [19,20]), extended to the upper floor (Ω,H2,P).
    Stated in §3 before (3.8) with a footnote saying the classical Wiener space proof extends 'easily'; no proof of the extension is given.
  • ad hoc to paper The estimates (2.4) and (2.5) hold: sup over t≤τ and over H-balls of |v_t|_H and ∇v_t are P-a.s. finite.
    These bounds are asserted to follow from H-analyticity and t-regularity, but their proof is not supplied; they carry the H-C^1 property of v^τ.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Strong solutions of SDE's with rough coefficients." pith.science (2026). https://pith.science/paper/GBQ6O5MY

@misc{pith2026250721592,
  author       = {Pith},
  title        = {Pith review of: Strong solutions of SDE's with rough coefficients},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GBQ6O5MY}},
  note         = {Machine review of arXiv:2507.21592}
}
read the original abstract

We give a proof of the strong existence and the regularity of stochastic differential equations driven by a Brownian motion and a measurable, Markovian drift without no regularity hypothesis except that the Girsanov exponential associated is in some L^{1+{\epsilon}}(\mu) for some fixed {\epsilon}>0 by using the techniques which are totally novel originating from the abstract Wiener space, in particular the solution is an H-C-regular map in the sense of the theory of Leonard Gross.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

21 extracted references · 20 canonical work pages

  1. [21]

    A. S. ¨Ust¨ unel and M. Zakai:Transformation of Measure on Wiener Space . Springer Verlag, 1999. A.S. ¨Ust¨ unel, 70, rue Bobillot, 75013 Paris, F rance or M¨ unir Nur. Sel¸ cuk Cadd. No.16/18, F enerbah¸ ce, Istanbul, T urkey , asustunel@gmail.com

    work page 1999

  2. [1]

    On the variational method for Euclidean quantum fields in infinite volume

    N. Barashkov and M. Gubinelli:“On the variational method for Euclidean quantum fields in infinite volume”. Prob. Math. Phys. 4 p. 761-801, 2023

    work page 2023

  3. [2]

    Uniqueness of solutions of stochastic differential equations

    A. M. Davie: “Uniqueness of solutions of stochastic differential equations”. International Mathematics Research Notices, 2007, 2007

    work page 2007

  4. [3]

    Dunford and J.T

    N. Dunford and J.T. Schwartz: Linear Operators, Vol. 2 , New York, Interscience, 1967

    work page 1967

  5. [4]

    Pathwise uniqueness and continuous dependence for SDEs with non-regular drift

    Ennio Fedrizzi and Franco Flandoli:“ Pathwise uniqueness and continuous dependence for SDEs with non-regular drift”. Stochastics: An International Journal of Probability and Stochastic Processes, 83(03):241–257, 2011

    work page 2011

  6. [5]

    Abstract Wiener spaces

    Leonard Gross:“Abstract Wiener spaces”. Proc. 4 th Berkley Sym. Math. Stat. Prob. 2, 31-42, 1965

    work page 1965

  7. [6]

    Variational calculus for diffusions

    K´ evin Hartmann:“Variational calculus for diffusions”. arXiv 1607.05488v3 [math PR] [Submitted on 19 Jul 2016 (v1), last revised 1 Dec 2016 (this version, v3)]

    work page Pith review arXiv 2016

  8. [7]

    Ikeda and S

    N. Ikeda and S. Watanabe: Stochastic Differential Equations and Diffusion Processes. North Holland, Amsterdam (Kodansha Ltd., Tokyo), 1981

    work page 1981

Show all 21 references
  1. [8]

    Strong solutions to stochastic equations with singular time dependent drift

    N.V. Krylov and M. R¨ ockner:“ Strong solutions to stochastic equations with singular time dependent drift”. Probab. Theory Relat. Fields 131 (2005), 154-196

    2005

  2. [9]

    Kuo: Gaussian Measures on Banach Spaces

    H.-H. Kuo: Gaussian Measures on Banach Spaces . lecture Notes in Math., Vol. 463, Springer, 1975. STRONG SOLUTIONS OF SDE WITH ROUGH COEFFICIENTS 7

    1975

  3. [10]

    Precise asymptotics of certain Wiener functionals

    S. Kusuoka and D.W. Stroock:“Precise asymptotics of certain Wiener functionals”. Jour. Func. Analysis, 99, p.1-74, 1991

    1991

  4. [11]

    Construction of strong solutions of SDE’s via Malliavin calculus

    T. Meyer-Brandis and F. Proske:“ Construction of strong solutions of SDE’s via Malliavin calculus”. Journal of Functional Analysis, 258(11):3922–3953, 2010

    2010

  5. [12]

    SDES with critical time dependent drifts:strong solutions

    M. R¨ ockner and G. Zhao:“SDES with critical time dependent drifts:strong solutions”. arXiv:2103.05803v5 [math.PR] 3 Jun 2025

    2025 arXiv

  6. [13]

    An example of stochastic differential equation having no strong solution

    B.S. Tsirelson: “An example of stochastic differential equation having no strong solution”. Theor. Prob. Appl. 20, p. 416-418, 1975

    1975

  7. [14]

    Representation of the distributions on Wiener space and stochastic calculus of variations

    A. S. ¨Ust¨ unel : “ Representation of the distributions on Wiener space and stochastic calculus of variations ”. Journal of Functional Analysis, Vol. 70,p. 126–139( 1987)

    1987

  8. [15]

    A. S. ¨Ust¨ unel:Introduction to Analysis on Wiener Space . Lecture Notes in Math. Vol. 1610. Springer, 1995

    1995

  9. [16]

    A. S. ¨Ust¨ unel:Analysis on Wiener Space and Applications . http://arxiv.org/abs/1003.1649, 2010

    2010 arXiv

  10. [17]

    A necessary and sufficient condition for invertibility of adapted perturbations of identity on Wiener space

    A. S. ¨Ust¨ unel:“A necessary and sufficient condition for invertibility of adapted perturbations of identity on Wiener space”. Comptes Rendus Math´ ematiques, Vol. 346, p. 897-900.2008

    2008

  11. [18]

    The construction of filtrations on abstract Wiener space

    A. S. ¨Ust¨ unel and M. Zakai: “The construction of filtrations on abstract Wiener space”. J. Funct. Anal.143 , p. 10–32, 1997

    1997

  12. [19]

    Entropy, invertibility and variational calculus of adapted shifts on Wiener space

    A. S. ¨Ust¨ unel : “Entropy, invertibility and variational calculus of adapted shifts on Wiener space”. Journal of Functional Analysis, Volume 257, Issue 11, Pages 3655-3689, 2009

    2009

  13. [20]

    Variational calculation of Laplace transforms via entropy on Wiener space and applications

    A. S. ¨Ust¨ unel: “Variational calculation of Laplace transforms via entropy on Wiener space and applications”. J. Funct. Anal. 267, no. 8, 3058–3083, 2014

    2014

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.