Pith. sign in

REVIEW 2 major objections 4 minor 2 cited by

Stochastic motions of the two-dimensional many-body delta-Bose gas, II: Many-$\delta$ motions

T0 review · 2 major / 4 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read For every N≥3, there are strong Markov processes of N planar particles that realize pairwise delta contact interactions at the path level and almost surely never bring three particles together before the terminal time.

desk verdict The many-δ construction is a real step for the 2D delta-Bose gas program; the soft spots are the unpublished companion [7] and a repairable jump-term slip in Section 4, not the gap the stress test claims. read the letter →

arxiv 2505.01704 v1 pith:TTLFUUPO submitted 2025-05-03 math.PR

Yu-Ting Chen This is my paper · ORCID
classification math.PR MSC 60J5560J6560H3081S40
keywords many-bodydelta-Bosegasstochasticmany-δmotionsno-triple-contactsno-simultaneous-contactsGirsanovtransformationsingulardifferentialequationMacdonaldfunctionsstrongMarkovprocesses
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 constructs stochastic many-δ motions: random motions of N point particles in the plane that realize, at the path level, the contact interactions of the two-dimensional many-body delta-Bose gas. For every N≥3, every set of coupling constants, and every weight system with at least two active pairs, the main theorem produces probability measures on particle paths up to a terminal time under which the particles obey an explicit singular stochastic differential equation whose drift is written in terms of the Macdonald functions $K_0$ and $K_1$. The construction also proves a pathwise no-triple-contacts property: before the terminal time, three particles almost surely never occupy the same point simultaneously. The proof proceeds by changing measure locally via Girsanov transformations of the stochastic one-δ motions from the companion preprint [7], concatenating the resulting processes at contact-creation times, and closing the argument with a theorem on the impossibility of simultaneous contacts. If the theorem is right, these are the path-level stochastic counterparts of multiple two-body δ potentials in two dimensions, and they provide the process ingredient for the Feynman–Kac-type formulas developed in the next paper of the series.

What carries the argument

The load-bearing object is the stochastic many-δ motion, defined as a strong Markov process satisfying the singular SDE (3.9)–(3.10) whose drift coefficients are written in terms of the Macdonald functions $K_0$ and $K_1$ and the weights $w$; before the first contact the drift is the logarithmic gradient of $\Psi(z)=\sum_j w_j K_0(\sqrt{\beta_j}\,|z_{j'}-z_j|)$. The construction uses three tools. First, local Girsanov transformations of the one-δ motions from [7], with Radon–Nikodym derivative processes whose logarithms are expanded by an approximate Itô formula into finite-variation terms involving the local time of the one-δ process at zero, a continuous local martingale, and its quadratic variation (Proposition 3.7). Second, weighted-average probability measures (Definition 3.10): before any contact, the law is a weighted average of the laws of stochastic one-δ motions, with weights determined by $w$; this is a multi-particle analogue of the exponential change of measure used to construct $\mathrm{BES}(0,\beta^\downarrow)$. Third, the no-simultaneous-contacts theorem (Theorem 3.4): a comparison argument showing that if finitely many nonnegative semimartingales with Bessel-type drifts and pairwise correlations bounded by $1$ are summed, the sum almost surely stays positive unless a single component hits zero; this yields no-triple-contacts. The pieces are assembled by the standard concatenation theorem for strong Markov processes.

What would settle it

Take N=3 with all three pair weights $w_j>0$ and initial positions with no coincidences. The theorem asserts that the first time any pair distance is zero and the first time a second pair distance is zero are almost surely distinct, i.e. $P(\exists t<T_\partial \text{ and distinct } j_1,j_2,j_3 \text{ with } Z^{j_1}_t=Z^{j_2}_t=Z^{j_3}_t)=0$. A concrete disproof would be to show, for some eligible initial configuration and weights, that this event has positive probability—for example, by constructing a positive-probability set of Brownian paths under (3.9) on which two different pair distances hit zero at the same time.

Watch

Extended reading notes

Core claim

The central claim is the existence, for all N≥3, of the claimed stochastic many-δ motions. For eligible initial configurations—those in which at most one active pair of particles is initially in contact—the paper defines a family P of probability measures on paths in $C^N \cup \{\partial\}$ under which the coordinate process is a time-homogeneous strong Markov process with an absorbing state $\partial$. Up to the terminal time $T_\partial$, the process solves the singular SDE (3.9)–(3.10), whose drift before contact is the logarithmic gradient of $\Psi(z)=\sum_j w_j K_0(\sqrt{\beta_j}\,|z_{j'}-z_j|)$, equivalently expressed through the Macdonald functions $K_0$ and $K_1$, and whose driving noises are independent planar Brownian motions obtained by Girsanov transformation. The no-triple-contacts property holds pathwise: the probability that three distinct particles occupy the same point at some time $t<T_\partial$ is zero. The proof builds two classes of strong Markov processes with lifetime—one defined by a weighted average of the laws of the stochastic one-δ motions before any contact, the other by local Girsanov transformations that close at the first contact time of a new pair—and concatenates them at contact-creation times.

Load-bearing premise

The construction imports the stochastic one-δ motions and their fine analytic properties—SDEs, transition densities, exponential moment bounds, and local-time growth—from the unpublished companion preprint [7], and every Girsanov transformation and local-time formula in this paper is applied to those processes, so an error in [7] would propagate into the main theorem.

Editorial extensions

If this is right

  • For any N≥3 and any eligible starting configuration, the SDE (3.9)–(3.10) admits a strong Markov solution up to the terminal time, giving a diffusion-process representation of multiple two-body δ interactions in two dimensions.
  • At each contact-creation time exactly one new pair begins to interact (no-triple-contacts), so the same two-step construction can be restarted from the new configuration, making the concatenation scheme well-defined.
  • When all active couplings coincide (w-homogeneous β), every contact-creation time is almost surely finite, so the inductive concatenation never stalls.
  • For {0,1}-valued weights, the stochastic many-δ motion is a probabilistic counterpart of the formal Hamiltonian with δ-potentials on the selected pairs, and it reduces to stochastic one-δ motions by change of measure.
  • The log-Radon–Nikodym semimartingale decompositions (Proposition 3.7) are the intended main tool for the Feynman–Kac-type formulas in the next paper [8].

Reading between the lines

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

  • If the weighted-average measures of Definition 3.10 can be shown to be true martingales rather than merely supermartingales up to the first contact, the phrase 'conditioned to attain contacts' would correspond to literal conditioning; the paper only establishes supermartingale and local martingale status, so this is an open strengthening.
  • The abstract no-simultaneous-contacts theorem may transfer to other singular interacting particle systems whose radial parts have Bessel-type drifts; checking its μ and correlation conditions is the entry point.
  • The paper leaves open whether the terminal time $T_\partial$ is almost surely infinite; if it can be finite with positive probability in inhomogeneous couplings, the absorbing state $\partial$ would be genuinely reached, and one would need to understand the boundary behaviour of the process there.
  • One practical consequence of pairwise-only contacts is that a simulation could alternate between free planar Brownian motion and two-body delta-contact reweighting, avoiding the simultaneous singularity of the N-body drift; this is not stated in the paper.
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

2 major / 4 minor

Summary. The paper constructs, for every integer N ≥ 3, families of probability measures on path space that make the coordinate process a time-homogeneous strong Markov process with an absorbing state ∂, solving the singular SDE (3.9) up to a terminal time T∂ and satisfying the no-triple-contacts (NTC) property. The construction has two ingredients: a first class of processes, defined for contact-free initial data as a weighted average of the one-δ laws from the companion paper [7] (Definition 3.10, Proposition 3.11), and a second class, defined by a local Girsanov transformation and handling initial conditions with one contact (Proposition 3.15). The two classes are concatenated using NTC, which is deduced from a general no-simultaneous-contacts theorem (Theorem 3.4). Section 5 supplies the required Itô-formula-type semimartingale decompositions for the logarithms of the Radon–Nikodým derivative processes.

Significance. If the main theorem is correct, this is a substantial advance: it provides path-level stochastic motions for the two-dimensional many-body delta-Bose gas, with explicit SDEs involving Macdonald functions, a pathwise proof of the no-triple-contacts phenomenon, and a clean Markovian framework for the Feynman–Kac-type formulas announced for [8]. The paper is also commendably candid about its open points, such as the explicit admission in Remark 3.3(1◦) that it is not known whether P(T∂ = ∞) = 1. However, two issues prevent me from accepting the paper in its present form: the proof of the key NTC theorem contains an unhandled jump term in the comparison argument, and the construction is heavily dependent on the unpublished companion preprint [7] for many load-bearing fine properties. The first issue is local and appears repairable; the second is a verification concern that should be addressed explicitly.

major comments (2)
  1. [Section 4, Step 4, Eq. (4.17)] The displayed equality in (4.17) is not valid for the càdlàg process Δ_ℓ^(k) = ρ_ℓ^(1),k − ρ_ℓ^(2). By Lemma 4.1 and (4.13), ρ^(2) and A have upward jumps whenever γ(ℓ−) < γ(ℓ), so Δ has downward jumps. For the convex regularizations φ_n, the correct Stieltjes change-of-variable formula contains the nonnegative jump sum Σ_s [φ_n(Δ_s) − φ_n(Δ_{s−}) − φ'_n(Δ_{s−}) ΔΔ_s], which is omitted in (4.17). Dropping a nonnegative jump term turns the asserted equality into a one-sided bound whose direction is not the one needed for the subsequent Gronwall argument. As written, the estimate (4.21), and hence the conclusion Δ ≤ 0, does not follow. Since Theorem 3.4 is the source of the NTC property used in Proposition 3.6, Lemma 3.19, Proposition 3.15(7◦), and the concatenation in Section 3.4, Theorem 3.1 is not fully established by the proof as written. The gap appears local and likely fixable — the monotonicity of the positive part at downward jumps should make the comparison go through once the jump part is treated separately — but the repair must be written out in full.
  2. [Throughout, especially Sections 3.1–3.3 and 5] The paper imports a large number of results from the unpublished companion preprint [7] without reproving them: the SDEs (3.18)–(3.20), transition densities [7, Theorem 2.1], integrability bounds [7, Propositions 4.2 and 4.5], the local-time growth [7, (3.2)], and the change-of-measure relation [7, (2.7)] used in Lemma 3.13. These inputs are load-bearing: the weighted-average definition (3.46), every Girsanov transformation, and the limiting arguments in Section 5 are applied to processes whose fine properties come from [7]. Because [7] is cited as “Chen, Y.-T. (2024+)” and is not part of the submitted manuscript, the main theorem is conditional on an external document that cannot be checked from the present text. The authors should either include the needed statements and proofs, or make [7] available in a citable and ref ereed form before final acceptance.
minor comments (4)
  1. [Remark 3.16(2◦)] Remark 3.16(2◦) refers to “Proposition 3.11 (5◦)”, but Proposition 3.11 has only properties (1◦)–(4◦); the intended reference is clearly Proposition 3.15(5◦).
  2. [Proposition 3.7(2◦)] The text says that {L_t^i} is the local time of the C-valued process {Z_t^i} at level 0; the normalization (3.37) makes clear that it is the local time of |Z_t^i| at level 0, and this should be stated explicitly.
  3. [Lemma 3.13 and following] The reference measure P^{(0)}_{z0} is used from Lemma 3.13 onward but is never defined in the paper; a definition or a precise statement of its role as the law of the relevant Brownian motion should be added.
  4. [Proof of Proposition 3.15(5◦)] The proof uses Hölder conjugates (p_0, q_0) with 1 < q_0 < 1 + 1/√2 without explaining the origin of this range; a brief justification would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: stochastic many-delta motions are explicitly constructed from one-delta laws via Girsanov and concatenation, with no prediction reducing to an input.

full rationale

The claimed theorem is constructive rather than derived from its own conclusion. For no-contact initial data, the law is explicitly defined in Definition 3.10, Eq. (3.46), as a weighted average of the one-delta laws with weights w and couplings beta as given inputs; the SDEs (3.9)-(3.10) are then verified under that measure via Girsanov and stochastic-integral arguments (Proposition 3.11(4)). The contact-initial-condition class is likewise defined by an explicit Girsanov density in (3.41) and (3.88), and the NTC/no-simultaneous-contacts property is proved by a comparison argument (Theorem 3.4, Section 4), not assumed. No parameter is fitted, and no claimed output is equivalent by construction to an input. The principal external dependency is the same-author companion preprint [7] for the one-delta motions and their fine properties, and [5] is used only as motivation; this is a load-bearing dependency and a verification risk rather than circularity, since [7] concerns the one-delta case and does not already contain the many-delta theorem claimed here. A possible gap in the proof of Theorem 3.4 around Eq. (4.17), where nonnegative jump terms are omitted in the change-of-variables step, is a correctness concern, not a circular step, and does not affect this circularity score.

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

No fitted free parameters: the weights w in R_+^{E_N} with at least two positive entries and the couplings beta in (0,infty)^{E_N} are model inputs parametrizing the pair interactions in the operator (1.4); Theorem 3.1 holds for all of them, so no value is fitted to data or chosen ad hoc. The paper adds no new physical entities; the absorbing state partial and terminal time T_delta are bookkeeping devices for terminating the concatenation. Its axioms are the foundation [7] (unpublished), the local-time normalization from [10], standard stochastic-calculus tools, and the no-triple-initial-contacts domain restriction (1.2). The axiomatic load is concentrated in [7], meaning the present contribution is conditional on the companion preprint; this is a dependency, not circularity.

assumptions (4)
  • domain assumption Existence and fine properties of the stochastic one-delta motions P^{beta_i downarrow, i} from [7]: the SDEs (3.18)-(3.20), transition densities ([7, Theorem 2.1]), integrability bounds ([7, Propositions 4.2 and 4.5]) and local-time growth ([7, (3.2)]).
    Every step of Section 3 transforms these processes: Definition 3.10 averages their laws, the second class changes measure from P^i_z0, and the log-sum Ito formulas in Proposition 3.7 are computed under P^i_z0. [7] is an unpublished companion preprint, so the present claims inherit its validity.
  • standard math The Markovian local time L^i of |Z^i| at 0 is normalized by E^i_0[int_0^infty e^{-q tau} dL^i_tau] = 1/log(1 + q/beta_i) (equation (3.37), from Donati-Martin and Yor [10]).
    This normalization fixes the local-time term (3.35) in the finite-variation part A; changing the normalization would change the drift in the SDEs. It is an input from prior literature, not derived here.
  • standard math Standard stochastic-calculus machinery: Bessel process facts, the Dambis-Dubins-Schwarz time change, the Yamada-Watanabe approximation functions, Erickson's skew-product extension [11], the Ichiba-Karatzas comparison principle [14], and Sharpe's concatenation theory for strong Markov processes [24].
    Used in Theorem 3.4 (Section 4) and in the concatenation step (Section 3.4); none of these are proven in the paper.
  • domain assumption Initial conditions satisfy (1.2): at most one pair (j',j) with w_j > 0 can start in contact, z_{j'} = z_j. Equivalently z0 lies in C^N_w-parallel or C^N_{i,notparallel,w\{wi}}-parallel (3.4).
    The construction processes contacts one pair at a time; initial triple contacts are outside the theorem's scope by design. The second class handles exactly one initial contact.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Stochastic motions of the two-dimensional many-body delta-Bose gas, II: Many-$\delta$ motions." pith.science (2026). https://pith.science/paper/TTLFUUPO

@misc{pith2026250501704,
  author       = {Pith},
  title        = {Pith review of: Stochastic motions of the two-dimensional many-body delta-Bose gas, II: Many-$\delta$ motions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TTLFUUPO}},
  note         = {Machine review of arXiv:2505.01704}
}
abstract

This paper is the second 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 construct and study the more general stochastic many-$\delta$ motions for $N$ particles. They have the interpretation of independent two-dimensional Brownian motions conditioned to attain the contact interactions that realize multiple two-body $\delta$-function potentials. For the construction, we transform the stochastic one-$\delta$ motions studied in [7] by Girsanov's theorem locally before a pair of particles with different initial conditions begins to contact each other. The strong Markov processes with lifetime thus obtained are concatenated by using the "no-triple-contacts" (NTC). This NTC phenomenon appears in the functional integral solutions of the two-dimensional many-body delta-Bose gas obtained earlier and is now proven at the pathwise level to a generalized degree.

Figures

Figures reproduced from arXiv: 2505.01704 by the authors.

Figure 1.1
Figure 1.1. The figure is a schematic diagram for the stochast [PITH_FULL_IMAGE:figures/full_fig_p003_1_1.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Continuous-time directed polymers to the Critical SHF via moments

    math.PR 2026-08 conditional novelty 6.0 of 10

    Rescaled point-to-point partition functions of continuous-time polymers with correlated Brownian noise converge in law to the critical Stochastic Heat Flow with an explicit strength parameter.

  2. Stochastic motions of the two-dimensional many-body delta-Bose gas, I: One-$\delta$ motions

    math.PR 2025-05 accept novelty 6.0 of 10

    For the 2D N-body delta-Bose gas with all but one delta interaction switched off, explicit transition densities, invariant measure, Harris recurrence and singular-drift SDEs are proved for the one-delta stochastic motion.

Reference graph

Works this paper leans on

32 extracted references · 19 canonical work pages · cited by 2 Pith papers

  1. [7]

    and Dol´ eans-Dade, C.A

    Ash, R.B. and Dol´ eans-Dade, C.A. (2000). Probability and Measure Theory . Second edition. Academic Press

    work page 2000

  2. [8]

    and Cancrini, N

    Bertini, L. and Cancrini, N. (1995). The stochastic heat equation: Feynman- Kac formula and intermittence. Journal of Statistical Physics 78 1377–1401. doi:10.1007/BF02180136

    doi:10.1007/bf02180136 1995

  3. [1]

    + 5∑ i=1 I5.7(i)t. (5.7) Here, the terms I5.7(i)t =I5.7(i) are defined as follows: I5.7(1) def= −wi(2βi) ∫ t 0 G1(R) 2R ⏐ ⏐ ⏐ ⏐ ⏐ R=2βi(εi+Ri s) × 2 H i 0(εi +Ris) ds, I5.7(2) def= −wi(2βi) ∫ t 0 G1(R) 2R ⏐ ⏐ ⏐ ⏐ ⏐ R=2βi(εi+Ri s) × 2|Z i s| H i 0(εi +Ris) dBi s, I5.7(3) def= 1 2wi(2βi)2 ∫ t 0 2G1(R) +RG0(R) 4R2 ⏐ ⏐ ⏐ ⏐ ⏐ R=2βi(εi+Ri s) × 4Ri s H i 0(εi +Ri...

    work page

  4. [2]

    First, the property that ( ˆK1/K0)(·) is increasing has been obtained in [6, Proposi- tion 4.12 (4 ◦)] by taking α = 0 there

    (5.51) Proof. First, the property that ( ˆK1/K0)(·) is increasing has been obtained in [6, Proposi- tion 4.12 (4 ◦)] by taking α = 0 there. Second, since K0 is decreasing, it suffices to show the convergence to zero in (5.50). In this case, we can just consi er the case that M = 1/2 since K0(·) is bounded away from zero on compacts in (0 , ∞). To prove the ...

    work page

  5. [3]

    This shows that the ri ght-hand side of (5.52) tends to zero as ε ց 0

    (5.57) Finally, we combine (5.53) and (5.57). This shows that the ri ght-hand side of (5.52) tends to zero as ε ց 0. We have proved (5.50) for M = 1/2. The proof is complete. ■ 5.5 End of the proof of Proposition 3.7 We show (3.33) first. Recall that Proposition 3.7 assumes z0 ∈ CN w\{wi} /parallelshort. By putting together (5.6), Lemmas 5.1–5.2, and Propo...

    work page

  6. [4]

    Since 0 < T <∞ is arbitrary, we obtain the property that with Pi z0-probability one, (3.33) holds for all t ≥ 0

    + 9∑ i=1 I5.8(i)t − 5∑ i=1 I5.7(i)t = log K β ,w 0 (0) wiK β ,i 0 (0) +Aβ ,w,i 0 (t) +N β ,w,i 0 (t) − 1 2 ⟨N β ,w,i 0 ,N β ,w,i 0 ⟩t, whereI5.8(i)t andI5.7(i)t on the right-hand side of the second equality are understood to use ε = εn. Since 0 < T <∞ is arbitrary, we obtain the property that with Pi z0-probability one, (3.33) holds for all t ≥ 0. The rem...

    work page

  7. [5]

    , Gesztesy, F

    Albeverio, S. , Gesztesy, F. , Høegh-Krohn, R. and Holden, H. (1987). Point interactions in two dimensions: Basic properties, approxi mations and applications to solid state physics. Journal f¨ ur die reine und angewandte Mathematik 380 87–107. doi:10.1515/crll.1987.380.87

    doi:10.1515/crll.1987.380.87 1987

  8. [6]

    , Høegh-Krohn, R

    Albeverio, S. , Høegh-Krohn, R. and Streit, L. (1977). Energy forms, Hamilto- nians, and distorted Brownian paths. Journal of Mathematical Physics 18 907–917. doi:10.1063/1.523359

    work page doi:10.1063/1.523359 1977

Show all 32 references
  1. [9]

    Chen, Y.-T. (2024). Delta-Bose gas from the viewpoint of the two-dimens ional stochastic heat equation. Annals of Probability 52 127–187. doi:10.1214/23-AOP1649

    2024 doi

  2. [10]

    Chen, Y.-T. (2022+). Two-dimensional delta-Bose gas: skew-product re lative motions. To appear in Annals of Applied Probability , available at arXiv:2207.06331

    2022 arXiv

  3. [11]

    Chen, Y.-T. (2024+). Stochastic motions of the two-dimensional many-b ody delta-Bose gas, I: One-δ motions. Preprint

    2024

  4. [12]

    Chen, Y.-T. (2024+). Stochastic motions of the two-dimensional many-b ody delta-Bose gas, III: Path integrals. Preprint

    2024

  5. [13]

    Chen, Y.-T. (2024+). Stochastic motions of the two-dimensional many-b ody delta-Bose gas, IV: Transformations of relative motions. Preprint

    2024

  6. [14]

    and Yor, M

    Donati-Martin, C. and Yor, M. (2006). Some explicit Krein representations of certain subordinators, including the Gamma process. Publications of the Research Institute for Mathematical Sciences 42 879–895. doi:10.2977/PRIMS/1166642190

    2006

  7. [15]

    Erickson, K.B. (1990). Continuous extensions of skew product diffusions. Probability Theory and Related Fields 85 73–89. doi:10.1007/BF01377630

    1990 doi

  8. [16]

    and Shepp, L.A

    Ezawa, H., Klauder, J.R. and Shepp, L.A. (1974). A path space picture for Feynman- Kac averages. Annals of Physics 88 588–620. doi:10.1016/0003-4916(74)90182-1

    1974 doi

  9. [17]

    F¨ollmer, H. (1972). The exit measure of a supermartingale. Zeitschrift f¨ ur Wahrschein- lichkeitstheorie und Verwandte Gebiete 21 154–166. doi:10.1007/BF00532472. 62

    1972 doi

  10. [18]

    and Karatzas, I

    Ichiba, T. and Karatzas, I. (2010). On collisions of Brownian particles. Annals of Ap- plied Probability 20 951–977. doi:10.1214/09-AAP641

    2010 doi

  11. [19]

    and W atanabe, S

    Itˆo, K. and W atanabe, S. (1965). Transformation of Markov processes by multiplica- tive functionals. Annales de l’Institut Fourier 15, 13–30. Available at Numdam at https://www.numdam.org/item/AIF 1965 15 1 13 0/

    1965

  12. [20]

    Kallenberg, O. (2002). Foundations of Modern Probability . Second edition. Springer- Verlag. doi:10.1007/978-1-4757-4015-8

    2002 doi

  13. [21]

    and Shreve, S

    Karatzas, I. and Shreve, S. (1998). Brownian Motion and Stochastic Calculus . Springer Science+Business Media New York. doi:10.1007/978-1-4612-0949-2

    1998 doi

  14. [22]

    Kardar, M. (1987). Replica Bethe ansatz studies of two-dimensional in ter- faces with quenched random impurities. Nuclear Physics B 290 582–602. doi:10.1016/0550-3213(87)90203-3

    1987 doi

  15. [23]

    Kardar, M. (2007). Statistical Physics of Particles . Cambridge University Press. doi:10.1017/CBO9780511815898

    2007 doi

  16. [24]

    and R¨ockner, M

    Krylov, N.V. and R¨ockner, M. (2005). Strong solutions of stochastic equations with singular time dependent drift. Probability Theory and Related Fields 131 154–196. doi:10.1007/s00440-004-0361-z

    2005 doi

  17. [25]

    Lebedev, N. N. (1972). Special Functions & Their Applications . Dover Publication

    1972

  18. [26]

    Protter, P.E. (2005). Stochastic Integration and Differential Equations . Second edition. Springer Berlin, Heidelberg. doi:10.1007/978-3-662-10061-5

    2005 doi

  19. [27]

    and Yor, M

    Revuz, D. and Yor, M. (1999). Continuous Martingales and Brownian Motion. 3rd edition. Springer-Verlag, Berlin, Heidelberg. doi:10.1007/978-3-662-06400-9

    1999 doi

  20. [28]

    Sharpe, M. (1988). General Theory of Markov Processes. Academic Press Inc

    1988

  21. [29]

    Streit, L. (1981). Energy Forms: Schroedinger Theory, Processes. Physics Reports 77 363–375. doi:10.1016/0370-1573(81)90084-3

    1981 doi

  22. [30]

    and V aradhan, S.R.S

    Stroock, D.W. and V aradhan, S.R.S. (2006). Multidimensional Diffusion Pro- cess. Classics in Mathematics. Reprint of the 1997 edition. Spri nger Verlag. doi:10.1007/3-540-28999-2

    2006 doi

  23. [31]

    Veretennikov, A.Yu. (1980). On strong solutions and explicit formulas for solu- tions of stochastic integral equations. Mathematics of the USSR-Sbornik 39 387–403. doi:10.1070/SM1981v039n03ABEH001522

    1980 doi

  24. [32]

    and W atanabe, S

    Yamada, T. and W atanabe, S. (1971). On the uniqueness of solutions of stochastic differential equations. Kyoto Journal of Mathematics 11 155–167. doi:10.1215/kjm/1250523691. 63

    1971

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.