Pith. sign in

REVIEW 3 major objections 3 minor 1 cited by

Full stochastic dynamics of a tracer in a dense single-file system

T0 review · 3 major / 3 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Every n-time cumulant of a tracer in a dense single-file system is a Gaussian integral over the conditional probabilities of one Brownian walker that must stay positive.

desk verdict Genuinely new connection between dense-SEP tracer correlations and Brownian excursions; the four-time formulas are plausible but need a uniformity argument or numerical check. read the letter →

arxiv 2505.21446 v1 pith:MQVVWCPN submitted 2025-05-27 cond-mat.stat-mech

Alexis Poncet , Aurélien Grabsch , Olivier Bénichou This is my paper
classification cond-mat.stat-mech MSC 82C2260K3560J6582C31 PACS 05.40.-a05.60.-k02.50.-r
keywords single-filediffusionsymmetricexclusionprocesstracerdynamicsmulti-timecorrelationsnon-MarkoviandenselimitquenchedversusannealedBrownianwalker
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 establishes that in a dense single-file system (Symmetric Exclusion Process with occupancy $\\rho \\to 1$), the full stochastic process of a tagged particle, although strongly non-Markovian and non-Gaussian, is generated by a single Markovian variable: a Brownian walker of diffusion constant $1/2$. All $n$-time annealed cumulants of the tracer are expressed as integrals over the probability that this walker stays positive at all intermediate times, and quenched cumulants follow from the same probabilities through logarithmic generating functions. The paper derives explicit four-time cumulants as combinations of square-root time differences, and extends the relation to two tracers, step density profiles, biased tracers, and finite observation times. This matters because it converts a strongly interacting many-body process into Gaussian integrals, giving concrete predictions for multi-time correlations and memory effects.

What carries the argument

The central object is the single-vacancy generating function, whose large-time limit is a Gaussian kernel for Brownian motion $K_\\tau(z',z)=e^{-(z'-z)^2/\\tau}/\\sqrt{\\pi\\tau}$. The argument proceeds by factorizing the tracer displacement into independent vacancy contributions, computing the single-vacancy propagator from first-passage statistics, and then taking the Brownian limit. The load-bearing identity is $P^+_n(z_0)=\\int_0^\\infty \\prod_{i=1}^n dz_i\, K_{\\tau_i}(z_i,z_{i-1})$, the probability that the walker started at $z_0<0$ stays positive at all observation times; every annealed cumulant is an integral of this quantity, and every quenched cumulant follows by differentiating the logarithm of the same generating function. For a biased tracer the Gaussian kernel is replaced by $K_{\\tau,s}$, and one sums over sign-change sequences of the walker.

What would settle it

Measure, by Monte Carlo simulation of the SEP at densities approaching 1, the rescaled four-time cumulant $\\lim_{\\rho\\to 1} \\langle X(t_1)X(t_2)X(t_3)X(t_4) \\rangle^{\\rm A}_c /(1-\\rho)$ for well-separated times and compare with the paper's explicit four-time formula; the predicted identity $\\kappa^{\\rm A}_4(t'_1,t'_2,t'_2,t'_2)=\\kappa^{\\rm A}_2(t'_1,t'_2)$ is the sharpest single check, and the annealed-versus-quenched large-$t_2$ behavior of the covariance (plateau versus decay) provides a second quantitative test.

Watch

Extended reading notes

Core claim

In the dense limit the tracer displacement generating function factorizes over independent vacancies, and at large times each vacancy is a Brownian walker $Z$ with diffusion constant $1/2$. The paper's central relation is $\\kappa^{\\rm A}_{2n}(t_1,\\ldots,t_{2n}) \\sim 2\\sqrt{2}\\int_{-\\infty}^{0} dz_0\\, P^+_{2n}(z_0)$, where $P^+_n(z_0)=\\int_0^\\infty \\prod_{i=1}^n dz_i\, K_{\\tau_i}(z_i,z_{i-1})$ is the probability that $Z$ stays positive at all times $t_i$, and the quenched cumulants are built from the same $P^+_n$ via logarithms, e.g. $\\kappa^{\\rm Q}_2 \\sim \\kappa^{\\rm A}_2 - 2\\sqrt{2}\\int_{-\\infty}^0 dz\\, P^+_1(z,t_1)P^+_1(z,t_2)$. All cumulants up to fourth order are computed explicitly: the annealed four-time cumulant is a sum of $\\sqrt{t_j - t_i}$ terms with coefficients involving $A(u)=(2/\\pi)\\arctan\\sqrt{u}$, and the quenched four-time cumulant combines $\\sqrt{t_j-t_i}$ and $\\sqrt{t_i+t_j}$ terms. The framework also gives the two-tracer covariance at separation $L$, the step-profile case, the biased tracer with kernel $K_{\\tau,s}(z',z)=[e^{-(z'-z)^2/\\tau}-\\nu' s\, e^{-(|z'|+|z|)^2/\\tau}]/\\sqrt{\\pi\\tau}$, and the all-times Laplace transform of Eq. (20).

Load-bearing premise

The load-bearing premise is that in the nearly full lattice vacancies are independent enough that the tracer displacement generating function factorizes into a product of single-vacancy contributions with errors only of order $(1-\\rho)^2$; the companion premise is that the large-time limit of each vacancy random walk is Brownian motion with diffusion constant $1/2$.

Editorial extensions

If this is right

  • All annealed cumulants of even order that involve the same set of $k$ observation times coincide, e.g. $\\kappa^{\\rm A}_4(t'_1,t'_2,t'_2,t'_2) = \\kappa^{\\rm A}_2(t'_1,t'_2)$, so fixing the time set fixes the whole $k$-time annealed statistic.
  • The annealed covariance is that of fractional Brownian motion with Hurst index $H=1/4$, and the same formula holds for any two-time cumulant $\\langle X(t_1)^p X(t_2)^q \\rangle_c$ in the dense limit.
  • Annealed and quenched initial conditions produce qualitatively different long-time behavior: with $t_1$ fixed and $t_2 \\to \\infty$ the quenched covariance decays to zero while the annealed covariance saturates at half the variance; quenched fourth-order cumulants are non-monotonic and can change sign.
  • A bias $s \\neq 0$ destroys the fractional-Brownian description of the annealed process; at $s = \\pm 1$ successive increments become uncorrelated, and the quenched covariance still decays to zero at large $t_2$.
  • A step density profile ($\\rho_+ \\neq \\rho_-$) leaves all even cumulants unchanged and forces the odd annealed cumulants to be proportional to the even ones with ratio $\\sigma = (\\rho_- - \\rho_+)/[2(1-\\rho)]$.

Reading between the lines

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

  • The time-set equality of annealed cumulants suggests an organizational principle beyond self-similarity: in the dense limit only the set of observation times, not which time is repeated, matters, pointing to the Brownian excursion as the fundamental variable rather than the tracer's own increments.
  • The annealed-versus-quenched contrast (covariance plateau at half the variance versus decay to zero) offers a concrete experimental handle: particle-tracking experiments that average over many starting configurations should see the plateau, while single long trajectories with fixed initial conditions should see the decay, providing a direct test of initial-condition memory.
  • The paper leaves implicit that the same reduction could be tested in other geometries, such as comblike structures or higher-dimensional lattices, where a single vacancy performs a different type of random walk and the $P^+_n$ integrals may reorganize into escape probabilities rather than staying-positive probabilities.
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.

Referee Report

3 major / 3 minor

Summary. The paper studies the multi-time statistics of a tagged particle (tracer) in the one-dimensional symmetric exclusion process in the limit where the density of particles tends to one. For this dense limit, the authors derive a general relation between all n-time annealed cumulants of the tracer position and the probability that a single Brownian walker starting at negative position stays positive at intermediate times; quenched cumulants are expressed through the same objects via logarithmic generating functions. They then evaluate these formulas explicitly up to fourth order, giving closed expressions involving arctangents of ratios of time differences for the four-time annealed and quenched cumulants, and they extend the approach to two tracers, step initial conditions, and a biased tracer. They also give Laplace-domain results for finite observation times. The derivation is based on the standard high-density vacancy picture: in the limit rho -> 1 the vacancies become independent and the tracer displacement decomposes into independent single-vacancy contributions; the single-vacancy propagator is computed exactly in Laplace space and then reduced by a scaling limit to Brownian motion. The known two-time covariance is reproduced as a check.

Significance. If correct, this is a significant step beyond the Gaussian description of single-file diffusion. The reduction of a non-Markovian, non-Gaussian tracer process to the conditional probabilities of a single Markovian walker is elegant and likely to be useful. The paper ships symbolic-computation notebooks, reproduces known two-time results, and provides explicit, falsifiable four-time predictions. The main caveats are the uncontrolled nature of the vacancy-factorization approximation for the new four-time content and a few technical inconsistencies in displayed formulas.

major comments (3)
  1. [SM Sec. I.C; main Eqs. (4)-(5)] The vacancy-independence factorization in SM Eqs. (S10)-(S11) is the only bridge from the interacting multi-vacancy SEP to the single-walker expressions (4)-(5), and it is stated to hold only up to O[(1-rho)^2]. The paper should make explicit that the results are for the sequential limit rho -> 1 first and then t_i - t_{i-1} -> infinity; for the simultaneous limit at fixed nonzero 1-rho, vacancy-vacancy encounters occur at a rate of order 1-rho, so the neglected terms are not controlled uniformly in time. I do not regard this as invalidating the formal statement, but because the new four-time content is exactly what is exposed to these corrections, I request either a uniformity argument for the factorization error or a direct numerical check of one four-time annealed and one quenched cumulant in the dense SEP. The known two-time checks do not exercise the beyond-Gaussian content.
  2. [End Matter, Eq. (20)] For n=1, Eq. (20) evaluates to +[u_1 u_2 ((2+u_1) s_2 + (2+u_2) s_1)]^{-1} with s_i = sqrt(u_i(2+u_i)), whereas Eq. (17) is the negative of this quantity. The general finite-time formula therefore appears to be missing a factor (-1)^n (or an equivalent sign convention). Please correct the formula or state the convention consistently.
  3. [Main text, Eq. (3)] The definition lambda_i = sum_{j=1}^i mu_j is inconsistent with the change of variables between position and increment variables used in SM Eq. (S46), where lambda_i = sum_{j=i}^n mu_j. With the printed definition, the exponent in Eq. (3) is not equivalent to SM Eq. (S47). Please fix the convention.
minor comments (3)
  1. [Beyond the large time limit] The displayed definition of K_n as a product of n increments makes K_1 the first moment of the displacement, which vanishes; however, Eq. (21) gives the single-time variance. Please clarify the indexing, for instance by defining K_1 as the variance of one increment or by stating that K_n denotes the appropriate nontrivial n-time correlation.
  2. [Main text, Eq. (4)-(10)] A short sentence stating explicitly that the large-time limit is taken after the dense limit rho -> 1 would remove the ambiguity about the order of limits and address a natural concern about uniformity in time.
  3. [Main text, Eq. (13)] The biased kernel in Eq. (13) is introduced without derivation; a brief explanation that it is the image-method solution for the Brownian limit of the biased vacancy walk would help the reader.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the dense-limit cumulants are derived from the microscopic vacancy dynamics, not assumed or fitted.

full rationale

The central claim, Eqs. (4)-(5) of the main text, is obtained by a chain of derivations in the Supplemental Material: the exact multi-vacancy decomposition (S9), the high-density factorization (S10)-(S11), the single-vacancy propagator computed from first-passage and no-touch probabilities of the vacancy random walk (S24)-(S32), and the large-time Brownian limit (S39)-(S43). Each of these steps is carried out in the paper and none of them takes the target n-time cumulant as an input. The single-vacancy propagator used in the Brownian limit is derived in SM Sec. I.D, and the Gaussian kernel (S43) follows from the asymptotic analysis of that propagator rather than being imposed as the desired answer. No parameter is fitted to any subset of the predicted cumulants; the final four-time expressions, Eqs. (9)-(10) and Tables I-II, are parameter-free functions of the times and the bias. The known two-time covariance (7)-(8) is recovered as a special case, which acts as a consistency check rather than as the source of the new content. The load-bearing factorization (S10), which neglects vacancy-vacancy correlations at O[(1-rho)^2], is an approximation whose uniformity in the large-time limit is not controlled; that is a possible correctness or rigor concern, not a circularity, because it does not make the derived cumulants equal to their inputs by construction. The self-citations in the paper provide standard single-vacancy quantities or prior single-time results, but the relevant propagators are rederived in the SM and the new multi-time cumulants are not set equal to those cited results. No self-definitional reduction, fitted-input-as-prediction, or self-citation chain forcing the result was found.

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

The paper introduces no free parameters; the density is taken to the limit ρ→1, and the bias s and step σ are model parameters, not fitted. The Brownian walker Z is a scaling limit of the vacancy process, not an invented entity. The main assumptions are the dense-limit factorization and the Brownian scaling limit, both standard and clearly stated.

assumptions (4)
  • domain assumption Independence of vacancies and factorization of the generating function in the ρ→1 limit (SM Eq. S10-S11).
    The paper assumes that each vacancy contributes independently to the tracer displacement, neglecting O[(1-ρ)^2] correlations. This is the basis for expressing the full generating function as a product of single-vacancy factors, which underlies all subsequent analysis.
  • domain assumption The large-time scaling limit in which a single vacancy's discrete random walk converges to Brownian motion with diffusion constant 1/2 (SM Eqs. S39-S43).
    The explicit formulas are derived in the limit of large time increments τ_i→∞ with z_i/√τ_i fixed. This Brownian approximation is standard for the SEP at long times but is an assumption about the scaling limit.
  • domain assumption In the single-vacancy system, the tracer moves only by one lattice spacing when the vacancy crosses the origin (SM Sec. I.D).
    This geometric constraint of the exclusion process is used to construct the propagator P^†(Y,τ,Z_1|Z_0) and to derive Eq. (S23).
  • standard math Standard analytic tools: Gaussian integrals, Owen T and S functions, and the relations in SM Sec. V.
    The evaluation of the multi-dimensional integrals relies on special functions and integral identities that are taken from the mathematical literature.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Full stochastic dynamics of a tracer in a dense single-file system." pith.science (2026). https://pith.science/paper/MQVVWCPN

@misc{pith2026250521446,
  author       = {Pith},
  title        = {Pith review of: Full stochastic dynamics of a tracer in a dense single-file system},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MQVVWCPN}},
  note         = {Machine review of arXiv:2505.21446}
}
abstract

Tracer diffusion in single-file systems, where particles are restricted to move on a line without passing each other, has been a fertile ground to investigate anomalous diffusion and strong memory effects. While the long-time behavior of such a tracer has been well studied, with a known subdiffusive dynamics and a Gaussian description for the rescaled position, the finer details of multi-time correlations remain poorly understood. This work focuses on the limit where almost all sites of a Symmetric Exclusion Process (SEP), a paradigmatic lattice model, are occupied. It extends beyond Gaussian descriptions and single-time statistics to address the multi-time correlation functions of the tracer in the SEP. In this dense limit, we present a general relation between all $n$-time correlations of the non-Markovian tracer position process and the conditional probabilities of a single Markovian random walker. Using this relation, we derive explicit expressions for the four-time correlations and further explore important extensions: multiple tracers, non-equilibrium situations, and finite observation times. Our results underscore significant memory effects, strong temporal correlations, and the influence of initial conditions on long-time dynamics.

Figures

Figures reproduced from arXiv: 2505.21446 by the authors.

Figure 1
Figure 1. FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3 [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

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

  1. An exactly solvable macroscopic fluctuation theory of single-file diffusion

    cond-mat.stat-mech 2026-07 conditional novelty 7.0 of 10

    For Brownian hard-rod single-file diffusion, exact large-deviation functions for tracer position and integrated current follow from a canonical mapping to point particles.

Reference graph

Works this paper leans on

56 extracted references · 49 canonical work pages · cited by 1 Pith paper

  1. [1]

    Spohn, Large Scale Dynamics of Interacting Particles (Springer Berlin Heidelberg, Berlin, Heidelberg, 1991)

    H. Spohn, Large Scale Dynamics of Interacting Particles (Springer Berlin Heidelberg, Berlin, Heidelberg, 1991)

    1991

  2. [2]

    ≡ κA even,2(t′ 1, t′ 2). There is no such equality for quenched cumulants, where the growing complexity of the expressions with the order of the cumulants comes from the non-linear, loga- rithmic, expression of the cumulant generating function with Pn derived above. Note however that the quenched cumulants can still be expressed in terms of the random wal...

    work page

  3. [3]

    Mallick, Physica A: Statistical Mechanics and its Ap- plications 418, 17 (2015)

    K. Mallick, Physica A: Statistical Mechanics and its Ap- plications 418, 17 (2015)

    work page 2015

  4. [4]

    K. Hahn, J. K¨ arger, and V. Kukla, Physical Review Letters 76, 2762 (1996)

    work page 1996

  5. [5]

    H¨ ofling and T

    F. H¨ ofling and T. Franosch, Reports on Progress in Physics 76, 046602 (2013)

    work page 2013

  6. [6]

    T. E. Harris, Journal of Applied Probability 2, 323 (1965), 3212197

    work page 1965

  7. [7]

    Arratia, The Annals of Probability 11, 362 (1983), 2243693

    R. Arratia, The Annals of Probability 11, 362 (1983), 2243693

    work page 1983

  8. [8]

    Peligrad and S

    M. Peligrad and S. Sethuraman, ALEA Lat. Am. J. Prob. Stat. 4, 245 (2008)

    work page 2008

Show all 56 references
  1. [9]

    Illien, O

    P. Illien, O. B´ enichou, C. Mej ´ ıa-Monasterio, G. Oshanin, and R. Voituriez, Physical Review Letters 111, 038102 (2013)

    2013

  2. [10]

    Hegde, S

    C. Hegde, S. Sabhapandit, and A. Dhar, Phys. Rev. Lett. 113, 120601 (2014)

    2014

  3. [11]

    P. L. Krapivsky, K. Mallick, and T. Sadhu, Phys. Rev. Lett. 113, 078101 (2014)

    2014

  4. [12]

    P. L. Krapivsky, K. Mallick, and T. Sadhu, J. Stat. Phys. 160, 885 (2015)

    2015

  5. [13]

    Imamura, K

    T. Imamura, K. Mallick, and T. Sasamoto, Physical Re- view Letters 118, 160601 (2017)

    2017

  6. [14]

    Mallick, H

    K. Mallick, H. Moriya, and T. Sasamoto, Physical Re- view Letters 129, 040601 (2022)

    2022

  7. [15]

    Grabsch, P

    A. Grabsch, P. Rizkallah, A. Poncet, P. Illien, and O. B´ enichou, Physical Review E107, 044131 (2023)

    2023

  8. [16]

    Grabsch and O

    A. Grabsch and O. B´ enichou, Physical Review Letters 132, 217101 (2024)

    2024

  9. [17]

    Poncet, O

    A. Poncet, O. B´ enichou, and P. Illien, Physical Review E 103, L040103 (2021). 6

    2021

  10. [18]

    Grabsch, A

    A. Grabsch, A. Poncet, P. Rizkallah, P. Illien, and O. B´ enichou, Sci. Adv.8, eabm5043 (2022)

    2022

  11. [19]

    Bettelheim, N

    E. Bettelheim, N. R. Smith, and B. Meerson, Phys. Rev. Lett. 128, 130602 (2022)

    2022

  12. [20]

    Krajenbrink and P

    A. Krajenbrink and P. L. Doussal, Physical Review Let- ters 127, 064101 (2021)

    2021

  13. [21]

    Integrability and ex- act large deviations of the weakly-asymmetric exclusion process,

    A. Krajenbrink and P. L. Doussal, “Integrability and ex- act large deviations of the weakly-asymmetric exclusion process,” (2025), arXiv:2505.12034 [cond-mat]

    2025 arXiv

  14. [22]

    P. L. Krapivsky, K. Mallick, and T. Sadhu, Journal of Statistical Mechanics: Theory and Experiment 2015, 09007 (2015)

    2015

  15. [23]

    Sadhu and B

    T. Sadhu and B. Derrida, Journal of Statistical Mechan- ics: Theory and Experiment 2015, 09008 (2015)

    2015

  16. [24]

    Leibovich and E

    N. Leibovich and E. Barkai, Physical Review E 88, 032107 (2013)

    2013

  17. [25]

    M. J. A. M. Brummelhuis and H. J. Hilhorst, Physica A: Statistical Mechanics and its Applications 156, 575 (1989)

    1989

  18. [26]

    Poncet, O

    A. Poncet, O. B´ enichou, V. D´ emery, and G. Oshanin, Physical Review E 97, 062119 (2018)

    2018

  19. [27]

    Poncet, A

    A. Poncet, A. Grabsch, O. B´ enichou, and P. Illien, Phys- ical Review E 105, 054139 (2022)

    2022

  20. [28]

    [26], we also investigate the dynamics of the vacancies in continuous time and obtain results be- yond the large time limit

    Following Ref. [26], we also investigate the dynamics of the vacancies in continuous time and obtain results be- yond the large time limit

  21. [29]

    See supplementary material at [url inserted by publisher]

  22. [30]

    Github repository with Mathematica notebooks at https://github.com/stykoo/TracerDenseSingleFile

  23. [31]

    D. B. Owen, Communications in Statistics - Simulation and Computation 9, 389 (1980)

    1980

  24. [32]

    Derrida and A

    B. Derrida and A. Gerschenfeld, Journal of Statistical Physics 136, 1 (2009)

    2009

  25. [33]

    Derrida and A

    B. Derrida and A. Gerschenfeld, Journal of Statistical Physics 137, 978 (2009)

    2009

  26. [34]

    S. F. Burlatsky, G. Oshanin, M. Moreau, and W. P. Reinhardt, Physical Review E 54, 3165 (1996)

    1996

  27. [35]

    Lizana, T

    L. Lizana, T. Ambj¨ ornsson, A. Taloni, E. Barkai, and M. A. Lomholt, Physical Review E 81, 051118 (2010)

    2010

  28. [36]

    Landim, S

    C. Landim, S. Olla, and S. B. Volchan, Communications in Mathematical Physics 192, 287 (1998/03/12/, 1998- 03-12)

    1998

  29. [37]

    B´ enichou, P

    O. B´ enichou, P. Illien, G. Oshanin, A. Sarracino, and R. Voituriez, Physical Review Letters 115, 220601 (2015)

    2015

  30. [38]

    Iomin, V

    A. Iomin, V. M´ endez, and W. Horsthemke, Fractional Dynamics in Comb-like Structures (WORLD SCIEN- TIFIC, 2018)

    2018

  31. [39]

    Venturelli, P

    D. Venturelli, P. Illien, A. Grabsch, and O. B´ enichou, Journal of Physics A: Mathematical and Theoretical 58, 215001 (2025), arXiv:2502.16951 [cond-mat]

    2025 arXiv

  32. [40]

    Universal scale-free decay of tracer-bath correlations in d-dimensional interacting particle systems,

    D. Venturelli, P. Illien, A. Grabsch, and O. B´ enichou, “Universal scale-free decay of tracer-bath correlations in d-dimensional interacting particle systems,” (2025), arXiv:2411.09326 [cond-mat]. End Matter C (4,e) 01 8s2 + u− (u− + u+a123 + u−a124 + u+a134) C (4,e) 02 u− (u...

    2025

  33. [41]

    Side comment: fractional Brownian motion 9

  34. [42]

    High-density CGF from the single-vacancy problem 10 D

    Annealed and quenched averages 10 C. High-density CGF from the single-vacancy problem 10 D. Propagator in the single-vacancy system 11 II. Unbiased tracer 12 A. Expression of the propagators 12 B. Annealed cumulants for increments at arbitrary times 13 C. Cumulant-generating f...

  35. [43]

    Annealed cumulants 15

  36. [44]

    Two-point covariance and step initial density profile 17 A

    Quenched cumulants 16 III. Two-point covariance and step initial density profile 17 A. Two-point covariance 17 B. Step initial condition 18 IV. Biased tracer 19 A. Expression of the propagators 19 B. Cumulant-generating functions at large times 20 C. Cumulants at large times f...

  37. [45]

    Annealed cumulants 22

  38. [46]

    Useful formulas 23 A

    Quenched cumulants 23 V. Useful formulas 23 A. Gaussian integrals 23 B. Integrals involving erfc 24 C. Integrals involving the Owen T function 24 D. Integrals involving the Owen S function 25 E. Useful relations 25 References 25 I. DENSE LIMIT OF THE SYMMETRIC EXCLUSION PROCES...

  39. [47]

    It is initially at the origin and we denote X(t) its position at time t

    Definitions Let us consider an tracer in the SEP. It is initially at the origin and we denote X(t) its position at time t. Our main goal is to characterize the stochastic process {X(t)}t≥0. This amounts to obtaining the n-time moments ⟨X(t1) . . . X(tn)⟩ for all times 0 < t1 ≤...

  40. [48]

    It is a Gaussian process (characterized only by its averages and covariances) that has zero mean and stationary increments

    Side comment: fractional Brownian motion Since the reader may not be familiar with continuous time stochastic processes, we provide a simple example: the fractional Brownian motion with Hurst index H. It is a Gaussian process (characterized only by its averages and covariances...

  41. [49]

    (S1) and (S3), we did not explicit what the averaging means

    Annealed and quenched averages In Eqs. (S1) and (S3), we did not explicit what the averaging means. There are two part to it: averaging over the initial conditions of the system, that is to say the initial positions of the particles in the SEP, and averaging over the stochasti...

  42. [50]

    As we are interested in returns of the vacancy at the origin, we will use the short-hand notation fZ(τ ) = f0,Z(τ )

    The probability fZ1,Z0 (τ ) of first passage at position Z1 and at time τ of the vacancy that was at initial position Z0 at time 0. As we are interested in returns of the vacancy at the origin, we will use the short-hand notation fZ(τ ) = f0,Z(τ )

  43. [51]

    The probability Π Z1,Z0 (τ ) that the vacancy moves from Z0 to Z1 during a time interval τ without touching the origin. Once we know these two quantities, the propagator P †(Y, τ, Z1|Z0) can be computed by partitioning on whether the vacancy touches the origin or not: P †(Y, τ...

  44. [52]

    (S48)-(S49) is that all the annealed cumulants of even order involving k times, κA even,k, are equal

    Annealed cumulants As explained in the main text, a striking consequence of Eqs. (S48)-(S49) is that all the annealed cumulants of even order involving k times, κA even,k, are equal. We give their expressions up to order k = 4, using the integrals of 16 section V. κA even,1(t)...

  45. [53]

    (S50)-(S51), using the integrals of section V

    Quenched cumulants The quenched cumulants of order 2 and 4 can be computed from Eqs. (S50)-(S51), using the integrals of section V. κQ 2 (t1, t2) ∼ 1√ 2π √t1 + t2 − √t2 − t1 , (S63) κQ 4 (t1, t2, t3, t4) ∼ 1 4 √ 2π X 1≤i<j≤4 D(4) ij p tj − ti + E(4) ij p ti + tj . (S64) The co...

  46. [55]

    (S123), one sees that all the even annealed cumulants involving k times, κeven,k are equal

    Annealed cumulants Looking at Eq. (S123), one sees that all the even annealed cumulants involving k times, κeven,k are equal. The same goes for the odd cumulants, κodd,k. The results can be summarized by the formulas κA even,k ∼ 22−k √ 2π X 0≤i<j≤k C (k,e) ij p tj − ti, (S130)...

  47. [56]

    atan α √ 1 + a2 β ! − atan α √ 1 + a2 β γp (1 + a2)α2 + β2 + γ2 !# + 1 β atan a + p α2 + β2 αβ

    Quenched cumulants The quenched cumulants, up to order n = 3, are given by κQ 1 ∼ s r 2t π (S149) κQ 2 ∼ 1√ 2π (1 + s2)√t1 + t2 − (1 − s2)√t2 − t1 − 2s2√t2 (S150) κQ 3 ∼ s 2 √ 2π   X 0≤i<j≤3 D(3) ij p tj − ti + X 1≤i<j≤3 E(3) ij p ti + tj   (S151) The coefficients for κQ 3...

    2008

  48. [4312]

    (S69) 17 We needed to introduce the quantities aijk , a′ ijk , bijkl , cijkl , dijkl , eijkl , e′ ijkl , fijkl , gijkl , hijkl are defined in terms of the function A given by Eq. (S62). aijk = A tj − ti tk − tj a′ ijk = A ti + tj tk − tj (S70) bijkl = A (tj − ti)(tl − tk) (tk ...

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.