A spectral approach to the narrow escape problem in two-dimensional domains

Louis Carillo; Thomas Normand; Tony Leli\`evre; Urbain Vaes

A spectral approach to the narrow escape problem in two-dimensional domains

T0 review · 0 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read For Brownian motion in a smooth 2D domain with N tiny exit windows, the paper proves that the exit rate and each window's exit probability have asymptotic expansions, to any order, in inverse powers of the logarithms of the window radii.

desk verdict Strong, careful paper delivering the first rigorous arbitrary-order expansions for the 2D narrow escape problem, including exit point laws; the only real flag is an imported regularity premise worth double-checking. read the letter →

arxiv 2608.05007 v1 pith:LLQUPU24 submitted 2026-08-05 math.AP math.PR

Louis Carillo , Tony Lelièvre , Thomas Normand , Urbain Vaes This is my paper

classification math.APmath.PR MSC 35P0535J2535C2060J65

keywords narrowescapeproblemquasi-stationarydistributionBrownianmotionDirichlet-NeumannLaplacianasymptoticexpansionquasimodeexitpointlogarithmicweights

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

The paper studies a Brownian particle in a smooth two-dimensional container whose boundary reflects the particle except for N small arcs (windows) through which it can escape. Assuming the particle starts in the quasi-stationary distribution, the exit time is exponential with rate given by the principal eigenvalue of the Laplacian with Dirichlet conditions on the windows and Neumann conditions elsewhere. The paper constructs a quasimode, an explicit approximate eigenfunction, in powers of K_k^ε = 1/|ln(ε_k/2)|, and proves that both the eigenvalue and the probabilities of leaving through each window admit asymptotic expansions to any prescribed order M. This gives, for general domains, a parameter-free spectral version of the narrow-escape asymptotics and, the authors state, the first mathematical expansion of the exit-point distribution.

What carries the argument

The quasimode is φ^ε(x) = 1 + Σ_{1≤|γ|≤M} (K^ε)^γ f_γ^ε(x), where each first-order f_k^ε is a sum of a singular logarithmic layer S_k^ε with arcsine weight w_k^ε(t) = χ_k^ε(t)(1−(t/T_{k,ε}^+)^2)^{-1/2} and a regular part R_k^ε solving a Neumann Poisson problem with compatibility constant a_k = π/|Ω|. Higher-order f_γ^ε are built recursively from v_γ and linear combinations of the f_k^ε, with coefficients b_{γ,j} chosen so that φ^ε is O(|K^ε|^M) uniformly on the Dirichlet windows. Green's formula then transfers the quasimode estimates to λ_0^ε, while, for the exit-point probabilities, auxiliary quasimodes with one window removed lead to a linear system whose leading matrix has an explicitly computed determinant p_M(K^ε); the rational-form expansion of the solution is what produces Theorem 4.6.

What would settle it

Take a disk with one exit window of radius ε and compute λ_0^ε numerically (for instance by finite elements or a spectral method) for several values of ε. If the difference between λ_0^ε and the paper's M-th order expansion does not decay like |K^ε|^{M+1}, equivalently like |ln ε|^{-(M+1)}, then the remainder claim is false. Similarly, simulate a two-window domain and compare the empirical exit frequency through window 1 with the predicted leading term K_1^ε/K^ε and its first corrections.

Watch

Extended reading notes

Core claim

For a C∞ bounded planar domain with N disjoint small exit arcs of radii ε_k, starting from the quasi-stationary distribution, the principal eigenvalue satisfies λ_0^ε = Σ_{1≤|γ|≤M} a_γ (K^ε)^γ + O(|K^ε|^{M+1}), with coefficients a_γ constructed recursively from solutions of linear elliptic problems and fixed by the requirement that the quasimode vanish approximately on the Dirichlet windows. The probability of exiting through window k_0 satisfies Y_{k_0}^ε = Σ_{q=0}^{M-1} Σ_{m=0}^q $Q^{{(k_0)}}$_{M,m,q}(K^ε)/p_M(K^ε)^{m+1} + O(|K^ε|^M), where p_M is an explicitly computed determinant, and the leading term is K_{k_0}^ε/K^ε. Because the quasimode is not in the domain of the operator but carries exact Neumann conditions and approximate Dirichlet conditions with controlled errors, the spectral computations are uniform in ε, yielding both the mean exit time (the inverse of the eigenvalue) and the full exit-point law in the small-window limit.

Load-bearing premise

Everything rests on a uniform regularity estimate for a family of auxiliary Neumann problems: the solutions must stay uniformly bounded in $W^{{1,p}}$ as the windows shrink, with constants independent of t and ε, and if that uniformity degenerates the remainder bounds collapse.

Editorial extensions

If this is right

Under the quasi-stationary distribution, the exit time is exponentially distributed with rate given by the M-th order expansion of λ_0^ε, with error O(|K^ε|^{M+1}).
The probability of leaving through window k has leading term K_k^ε/K^ε, so among windows with comparable logarithmic sizes the exit odds are proportional to the inverse-log weights 1/|ln(ε_k/2)|.
The exit-point expansion is obtained for general smooth two-dimensional domains, not only disks, and to arbitrary order, extending the previously known first-order mean-exit-time asymptotics.
The coefficient recursion, driven by the local geometry at the window centers encoded in the functions f_j^0(x^{(k)}) and v_γ, provides a concrete algorithm for computing higher-order corrections to both the rate and the exit probabilities.

Reading between the lines

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

A direct Monte Carlo simulation of reflected Brownian motion in a disk or rectangle with two windows could test the leading prediction P_k ≈ K_k^ε/K^ε, and the paper's expansion predicts the next-order corrections that such a simulation could look for.
The construction suggests that the narrow-escape problem inherits an Eyring-Kramers-like structure for entropic metastability: the exponential rate is controlled by geometrically computable coefficients (starting with π/|Ω|) rather than by an energy barrier, and the higher-order coefficients encode boundary curvature and inter-window distances.
The determinant p_M(K^ε) organizing the exit-probability expansion points to an algebraic structure in inverse-logarithm variables that may also appear in related problems such as narrow capture or escape in tubes, where similar spectral quasimodes are used.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Bottom line: this is a strong, careful paper that delivers the first rigorous arbitrary-order expansions for the narrow escape problem in general two-dimensional domains, including the per-window exit probabilities. It deserves a serious referee and, assuming the one flagged premise holds up, acceptance with minor revisions.

What's new: prior work [23] covered the disk with interior holes; here the windows are actual boundary arcs on a smooth domain, the quasimode is deliberately not in the domain of the operator, and the main theorems give expansions to arbitrary order for both the principal eigenvalue (mean exit time) and the exit point distribution. The construction is recursive and messy—there's no closed form for the higher coefficients, and the authors are honest that it's a recipe—but the algebra is internally consistent. I spot-checked Lemma 2.3's flux, the exponents in Lemmas 2.6-2.7, and the determinant structure in Lemma 4.4: all coherent. The leading-order formulas match the physics (exit rate ~ sum 1/|ln ε_k|, exit probabilities ~ 1/|ln ε_k|), and no fitted parameters appear anywhere; the coefficients are fixed by compatibility and cancellation conditions.

The real soft spot is Proposition 2.4. The entire uniformity of the quasimode boundary cancellation rests on the imported W^{1,p} bound from [3] being uniform in t and ε. I can't see a degeneration: Lemma 2.2 gives N_k uniformly in L∞, so standard Neumann theory should supply it, but this is the one premise I'd want a referee to verify carefully rather than take on faith. The other issue is a sign slip in Lemma 2.3: the flux I(t) actually computes to +π, not -π, as a quick check of (2.24)-(2.28) shows. The conclusion a_k = π/|Ω| is still correct because the compatibility condition involves -I(t), so this is a typographical defect, not a load-bearing error, but it should be fixed. The self-citation [6] for QSD facts is unverifiable from here, but Appendix A gives the PDE proof and the statements are standard.

Who this is for: anyone in PDE/spectral theory or applied probability working on metastability, narrow escape, or small-hole asymptotics. It's not light reading, but it's honest, self-contained, and the main claims are concrete. I'd bring it to a reading group and cite it. Send it to review.

Referee Report

0 major / 4 minor

Summary. The paper develops a spectral quasimode approach to the narrow escape problem for a Brownian motion in a smooth two-dimensional domain with reflecting boundary, except for N small disjoint Dirichlet arcs on the boundary. The main objects are the principal eigenvalue lambda_0^epsilon of the mixed Laplacian and the boundary normal derivative of the principal eigenfunction, which give respectively the exponential rate of the exit time and the exit-point distribution under the quasi-stationary distribution. The authors construct a quasimode with explicitly determined coefficients and prove, to any prescribed order M, expansions of lambda_0^epsilon in powers of K_k^epsilon = -1/ln(epsilon_k/2), and of the per-window exit probabilities Y_k^epsilon as rational functions of a determinant p_M(K^epsilon). The leading term of the exit-point expansion is K_{k0}^epsilon / K^epsilon, recovering the expected geometric leading behavior.

Significance. If the results are correct, this is a substantial contribution: it gives the first rigorous arbitrary-order asymptotic expansion of the exit-point distribution in a general smooth two-dimensional domain, in addition to the mean exit time, and it extends the authors' previous disk result to the natural setting of exits through boundary arcs. The construction is genuinely quasimodal rather than fitted: the coefficients a_gamma, b_{gamma,k}, and K_k^epsilon are determined by compatibility, recursion, and boundary cancellation, and the expansions are then derived. The paper also provides explicit recursive formulas, a determinant lower bound, and a self-contained identification of the exit measure as a probability measure. The main estimates are spot-checkable and internally consistent. The only reservations are local presentation issues and the need to make the imported uniform regularity premise fully transparent.

minor comments (4)

[Lemma 2.3] In the proof of Lemma 2.3, after (2.28) one has I2(t) = -pi + O(sqrt(delta)), so with I1(t)=0 and I3(t)=O(delta), equation (2.24) gives I(t) = 0 - (-pi) + O(delta) = +pi + O(sqrt(delta)), not -pi. The stated conclusion a_k = pi/|Omega| is nevertheless correct because the compatibility condition yields a_k|Omega| = I(t); the displayed value 'I(t) = -pi' and the sentence following (2.29) should be corrected.
[Lemma 2.6] In the proof of Lemma 2.6, the computation of ||tilde w_k^epsilon||_1 silently replaces the interval [T^-_{k,epsilon}, T^+_{k,epsilon}] by [-T^+_{k,epsilon}, T^+_{k,epsilon}]. Using only (2.1)-(2.2), the missing piece is O(epsilon_k^{3/2}), not O(|epsilon|^2), because tilde w_k^epsilon is integrably singular at the endpoint. The later statement ||w_k^epsilon||_1 = pi epsilon_k (1+O(|epsilon|^{1/2})) is unaffected, but the displayed O(|epsilon|^2) estimate should be corrected or justified.
[Proposition 2.4] The uniform W^{1,p} bound for tilde R_k(·,t) is imported from [3, Proposition 2.3] and is load-bearing: it feeds Corollary 2.10 and hence the boundary cancellation of the quasimode in Lemmas 2.12 and 2.13. Since the bound is asserted uniformly in t and epsilon, please include the precise statement of the imported result or a short proof that the constant depends only on Omega, p, and the uniform L^infty bound of N_k supplied by Lemma 2.2. I see no indication that the bound degenerates, but the uniformity should be made transparent.
[Section 2.1] The symbol O(|K^epsilon|^infty) is used repeatedly (for example in (4.7) and in the proof of Theorem 4.6) but is never defined. A short definition, e.g., bounded by C_m |K^epsilon|^m for every m, would improve readability.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the quasimode expansions are derived from coefficients fixed by compatibility, recursion, and boundary cancellation, not fitted to the target eigenvalues or exit probabilities.

full rationale

The central derivation is self-contained. The expansion variable K^ε_k is fixed by boundary cancellation at (2.43), the first-order coefficients a_k by the Neumann compatibility condition in Lemma 2.3, and the higher-order coefficients (b_{β,j}, a_γ) by the explicit recursion (2.46) and formula (2.20) in Lemma 2.13. None of these coefficients is fitted to λ_0^ε or Y^ε_k. Theorem 3.3 derives the λ_0^ε expansion from Proposition 2.14(d,e) via Green's formula, and Theorem 4.6 derives the Y^ε_k expansion by solving the independently constructed linear system A^ε_M Y^ε = B^ε_M of Lemma 4.3. The self-citations to the authors' prior works [6] and [23] provide context and probabilistic QSD background, but the spectral expansions and the exit-point system do not reduce to those papers. The external regularity input [3, Proposition 2.3] used in Proposition 2.4 is a real analytic premise, not a self-citation; its uniformity is worth checking but that is a correctness concern, not circularity. The apparent sign inconsistency in the proof of Lemma 2.3 (which would give I = +π rather than −π) does not change the conclusion a_k = π/|Ω| and is a typographical defect, not a circular step. Overall, no prediction is equivalent by construction to a fitted input, so the circularity score is low.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The paper introduces no free parameters and no new physical entities. Its input assumptions are standard mathematical background (elliptic regularity, spectral theory, QSD theory) plus the geometric regularity of the domain and windows. The only tunable object is the truncation order M, which is an analysis parameter, not a fitted constant.

assumptions (5)

standard math Elliptic regularity: the mixed boundary value problem (2.23) with L^infty Neumann data has a unique W^{1,p} solution with uniform bounds (Proposition 2.4, citing [3, Proposition 2.3]).
Used throughout Section 2 to define the regular parts R_k^epsilon and control their Hoelder continuity on the exit windows.
standard math Spectral theory of L^epsilon: L^epsilon has compact resolvent, discrete spectrum, simple first eigenvalue lambda_0^epsilon > 0 with sign-definite eigenfunction (Section 1.2, [16, Theorem 8.38], [23, Section 2]).
Justifies the QSD characterization (1.4)-(1.5) and the normalization (3.4).
standard math Quasi-stationary distribution facts: existence, uniqueness, convergence of the conditioned process, exponential law and independence of tau and X_tau (Section 1.1, [6], [7], [10]).
Motivates and defines the quantities lambda_0^epsilon and Y_k^epsilon; Appendix A re-derives the exit law from the PDE.
domain assumption C^infty smoothness of the domain and exit-window parametrization with |T_{k,epsilon}^+- epsilon_k| = O(|epsilon|^2) and |T_{k,epsilon}^-| <= T_{k,epsilon}^+ (Section 2.1, Equations (2.1)-(2.2)).
Defines the geometric setting; smoothness is used in Taylor expansions for S_k^epsilon and N_k, and the symmetric-window condition enters Lemmas 2.6 and 2.7.
domain assumption The process starts from the quasi-stationary distribution nu_0^epsilon (Section 1.1).
The theorems are stated for this initial law; for arbitrary initial conditions the exit time is not exponential and the expansions do not apply.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A spectral approach to the narrow escape problem in two-dimensional domains." pith.science (2026). https://pith.science/paper/LLQUPU24

@misc{pith2026260805007,
  author       = {Pith},
  title        = {Pith review of: A spectral approach to the narrow escape problem in two-dimensional domains},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LLQUPU24}},
  note         = {Machine review of arXiv:2608.05007}
}

read the original abstract

We study the law of the exit time and exit point of a Brownian motion in a two-dimensional domain with reflecting boundary conditions, except on small disjoint exit windows through which the stochastic process can escape the domain. In the limit of infinitely small exit windows, it is natural to assume that the process starts from the quasi-stationary distribution. In this setting, we obtain a precise description of the exit event.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

37 extracted references · 36 canonical work pages

[23]

Leli` evre, M

T. Leli` evre, M. Rachid, and G. Stoltz,A spectral approach to the narrow escape problem in the disk, 2024. https://arxiv.org/abs/2401.06903

arXiv 2024
[3]

J. Aramaki,Existence and Regularity for the Neumann Problem to the Poisson Equation and an Application to the Maxwell-Stokes Type Equation, Communications in Mathematical Analysis21(2018), no. 1, 54 –66

work page 2018
[6]

Carillo, T

L. Carillo, T. Leli` evre, T. Normand, G. Stoltz, and U. Vaes,The narrow escape problem in arbitrary dimension, 2026

work page 2026
[1]

Ammari, K

H. Ammari, K. Kalimeris, H. Kang, and H. Lee,Layer potential techniques for the narrow escape problem, J. Math. Pures Appl. (9)97(2012), no. 1, 66–84. MR2863765

work page 2012
[2]

Ammari, H

H. Ammari, H. Kang, and H. Lee,Layer potential techniques in spectral analysis, Mathematical Surveys and Mono- graphs, vol. 153, American Mathematical Society, Providence, RI, 2009. MR2488135

work page 2009
[4]

B´ enichou and R

O. B´ enichou and R. Voituriez,Narrow-escape time problem: Time needed for a particle to exit a confining domain through a small window, Phys. Rev. Lett.100(2008), 168105

work page 2008
[5]

Br´ ezis,Functional analysis, sobolev spaces and partial differential equations, Vol

H. Br´ ezis,Functional analysis, sobolev spaces and partial differential equations, Vol. 2, Springer, 2011

work page 2011
[7]

Champagnat and D

N. Champagnat and D. Villemonais,General criteria for the study of quasi-stationarity, Electron. J. Probab.28(2023), no. 22, 1–84

work page 2023

Show all 37 references

[8]

Chen and A

X. Chen and A. Friedman,Asymptotic analysis for the narrow escape problem, SIAM J. Math. Anal.43(2011), no. 6, 2542–2563 (English)

2011
[9]

A. F. Cheviakov, M. J. Ward, and R. Straube,An asymptotic analysis of the mean first passage time for narrow escape problems. II: The sphere, Multiscale Model. Simul.8(2010), no. 3, 836–870 (English)

2010
[10]

Collet, S

P. Collet, S. Mart´ ınez, and J. San Mart´ ın,Quasi-stationary distributions: Markov chains, diffusions and dynamical systems, Probability and Its Applications, vol. 1, Springer, 2013

2013
[11]

Di Ges` u, T

G. Di Ges` u, T. Leli` evre, D. Le Peutrec, and B. Nectoux,Jump Markov models and transition state theory: The quasi-stationary distribution, Faraday discussions195(2016), 469–495

2016
[12]

L. C. Evans,Partial differential equations, American Mathematical Society, Providence, R.I., 2010

2010
[13]

I Freidlin and A

M. I Freidlin and A. D Wentzell,Diffusion processes on graphs and the averaging principle, The Annals of probability (1993), 2215–2245

1993
[14]

M. I. Freidlin,Functional integration and partial differential equations, Annals of mathematics studies, Princeton University Press, Princeton (N.J.), 1985 (eng)

1985
[15]

G. D. Ges` u, T. Leli` evre, D. L. Peutrec, and B. Nectoux,Sharp asymptotics of the first exit point density, Ann. PDE 5(2019), no. 1, 5. 35

2019
[16]

Gilbarg and N

D. Gilbarg and N. S. Trudinger,Elliptic partial differential equations of second order, Classics in Mathematics, Springer- Verlag, Berlin, 2001

2001
[17]

S Grebenkov, R

D. S Grebenkov, R. Metzler, and G. Oshanin,Full distribution of first exit times in the narrow escape problem, New Journal of Physics21(2019), no. 12, 122001

2019
[18]

V Grigoriev, Y

I. V Grigoriev, Y. A Makhnovskii, A. M Berezhkovskii, and V. Y. Zitserman,Kinetics of escape through a small hole, The Journal of chemical physics116(2002), no. 22, 9574–9577

2002
[19]

Holcman and Z

D. Holcman and Z. Schuss,Escape through a small opening: Receptor trafficking in a synaptic membrane, J. Stat. Phys.117(2004), no. 5-6, 975–1014 (English)

2004
[20]

Hsu,Asymptotic analysis on narrow tubes: narrow escape problems and diffusion processes, 2025

W.-T. Hsu,Asymptotic analysis on narrow tubes: narrow escape problems and diffusion processes, 2025. https://arxiv.org/abs/2508.15060v2

2025 arXiv
[21]

https://arxiv.org/abs/2511.21548v2

,Metastability of diffusion processes in narrow tubes, 2025. https://arxiv.org/abs/2511.21548v2

2025
[22]

Leli` evre, D

T. Leli` evre, D. Le Peutrec, and B. Nectoux,Eyring–Kramers exit rates for the overdamped Langevin dynamics: The case with saddle points on the boundary, Journal de l’Ecole Polytechnique – Math´ ematiques12(2025), 881–982

2025
[24]

J. L. Lions and E. Magenes,Non-homogeneous boundary value problems and applications: Vol. 1, Vol. 1, Springer Science & Business Media, 2012

2012
[25]

Lions and A.-S

P.-L. Lions and A.-S. Sznitman,Stochastic differential equations with reflecting boundary conditions, Communications on Pure and Applied Mathematics37(1984), no. 4, 511–537

1984
[26]

N. I. Muskhelishvili,Singular Integral Equations : boundary problems of functions theory and their applications to mathematical physics, P. Noordhoff, 1953

1953
[27]

Nursultanov, J

M. Nursultanov, J. C Tzou, and L. Tzou,On the mean first arrival time of brownian particles on riemannian manifolds, Journal de Math´ ematiques Pures et Appliqu´ ees150(2021), 202–240

2021
[28]

Pillay, M

S. Pillay, M. J. Ward, A. Peirce, and T. Kolokolnikov,An asymptotic analysis of the mean first passage time for narrow escape problems. I: Two-dimensional domains, Multiscale Model. Simul.8(2010), no. 3, 803–835 (English)

2010
[29]

Rudin,Real and complex analysis(1974)

W. Rudin,Real and complex analysis(1974)

1974
[30]

Schuss, A

Z. Schuss, A. Singer, and D. Holcman,The narrow escape problem for diffusion in cellular microdomains, Proceedings of the National Academy of Sciences104(2007), no. 41, 16098–16103

2007
[31]

Singer, Z

A. Singer, Z. Schuss, and D. Holcman,Narrow escape, Part II: The circular disk, J. Stat. Phys.122(2006), no. 3, 465–489 (English)

2006
[32]

Singer, Z

A. Singer, Z. Schuss, D. Holcman, and R. S. Eisenberg,Narrow escape. I, J. Stat. Phys.122(2006), no. 3, 437–463 (English)

2006
[33]

Singer, Z Schuss, and D

A. Singer, Z Schuss, and D. Holcman,Narrow escape and leakage of brownian particles, Physical Review E78(2008), no. 5, 051111

2008
[34]

S Strichartz,A guide to distribution theory and fourier transforms, World Scientific Publishing Company, 2003

R. S Strichartz,A guide to distribution theory and fourier transforms, World Scientific Publishing Company, 2003

2003
[35]

W Stroock and S

D. W Stroock and S. S. Varadhan,Diffusion processes with boundary conditions, Communications on Pure and Applied Mathematics24(1971), no. 2, 147–225

1971
[36]

F. G. Tricomi,Integral equations, [4th ed.], Pure and applied mathematics, Interscience, New York, 1967 (eng)

1967
[37]

Watanabe,On stochastic differential equations for multi-dimensional diffusion processes with boundary conditions, Journal of Mathematics of Kyoto University11(1971), no

S. Watanabe,On stochastic differential equations for multi-dimensional diffusion processes with boundary conditions, Journal of Mathematics of Kyoto University11(1971), no. 1, 169–180

1971

Pith tools

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

[1] [23]

Leli` evre, M

T. Leli` evre, M. Rachid, and G. Stoltz,A spectral approach to the narrow escape problem in the disk, 2024. https://arxiv.org/abs/2401.06903

arXiv 2024

[2] [3]

J. Aramaki,Existence and Regularity for the Neumann Problem to the Poisson Equation and an Application to the Maxwell-Stokes Type Equation, Communications in Mathematical Analysis21(2018), no. 1, 54 –66

work page 2018

[3] [6]

Carillo, T

L. Carillo, T. Leli` evre, T. Normand, G. Stoltz, and U. Vaes,The narrow escape problem in arbitrary dimension, 2026

work page 2026

[4] [1]

Ammari, K

H. Ammari, K. Kalimeris, H. Kang, and H. Lee,Layer potential techniques for the narrow escape problem, J. Math. Pures Appl. (9)97(2012), no. 1, 66–84. MR2863765

work page 2012

[5] [2]

Ammari, H

H. Ammari, H. Kang, and H. Lee,Layer potential techniques in spectral analysis, Mathematical Surveys and Mono- graphs, vol. 153, American Mathematical Society, Providence, RI, 2009. MR2488135

work page 2009

[6] [4]

B´ enichou and R

O. B´ enichou and R. Voituriez,Narrow-escape time problem: Time needed for a particle to exit a confining domain through a small window, Phys. Rev. Lett.100(2008), 168105

work page 2008

[7] [5]

Br´ ezis,Functional analysis, sobolev spaces and partial differential equations, Vol

H. Br´ ezis,Functional analysis, sobolev spaces and partial differential equations, Vol. 2, Springer, 2011

work page 2011

[8] [7]

Champagnat and D

N. Champagnat and D. Villemonais,General criteria for the study of quasi-stationarity, Electron. J. Probab.28(2023), no. 22, 1–84

work page 2023