{"id":"8f9e107a-ac6c-4fdf-9b39-f54fbcf77a04","arxiv_id":"2505.01533","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The extremes of many diffusing particles in a random environment obey KPZ statistics whose scaling exponent is set by the lowest randomly fluctuating moment of the environment.","lead":"This paper derives the extreme value statistics of many particles diffusing in a random space-time environment, showing they follow the Kardar-Parisi-Zhang universality class. It predicts that the scaling of these extremes is controlled by the lowest random moment of the environment, forming a hierarchy of universality classes.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The paper's own text admits the extreme-first-passage-time hierarchy fails for m>3, so the central claim of an infinite super-universal hierarchy for both extreme observables is overstated as stated.","rationale":"The reader identified the unproven independence of environmental and sampling fluctuations as the weakest assumption. That is a real gap, and Eqs. 12–13 depend on it, but the paper's own admission that the first-passage-time predictions break down for m>3 is more directly load-bearing: it is an internally acknowledged failure of the central claim as formulated for an infinite hierarchy of both extreme observables. The proposed m=4 FPT simulation would settle whether Eq. 7 should be restricted to m≤3. The location hierarchy and the m≤3 FPT results retain value, so the paper remains conditionally acceptable rather than rejected; the reader's verdict does not need to change, but the abstract and conclusion should be revised to match the paper's stated limitations. I therefore mark partial agreement: the reader's concern is valid but not the decisive one.","tokens_in":20922,"tokens_out":5934,"duration_ms":65783,"concrete_test":"Simulate the extreme first passage time for an m=4 environment, e.g. the random three-step distribution with k=10, at several barrier positions L and N≈10^28 in the stated asymptotic regime L^{4m-2}≫λ_ext^2 ln(N)^{4m-1}. Measure Varν(Env_L^N) and compare its L-dependence with Eq. 7, which predicts ∼L^{-3}. If the measured exponent differs or the variance no longer collapses, Eq. 7 must carry an explicit m≤3 restriction, and the abstract/conclusion should be revised to claim the infinite hierarchy only for the extreme location.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The Numerical Results section states: \"For the extreme first passage time and m>3, my theoretical predictions break down. This is because my predictions for the extreme first passage time are derived by translating the KPZ fluctuations at R(t)∝t^{(4m-1)/4m} to the bulk regime where R(t)∝O(1).\" This is an explicit, internally acknowledged limitation, yet Eq. 7 is presented without an m≤3 restriction, the abstract claims the scalings of \"the extreme location and extreme first passage time\" depend on the moments, and the conclusion asserts an infinite hierarchy of universality classes. Thus the strongest form of the central claim—Eqs. 6 and 7 giving an infinite super-universal hierarchy for both extremes—is not supported by the paper's own results. At minimum, the FPT hierarchy is established only for m=1,2,3, and the non-backtracking approximation underlying the FPT derivation is unquantified and fails in the bulk regime for m>3. Separately, for m>2 the environmental variance in Eq. 6 decays to zero as t→∞, so the higher \"universality classes\" are transient scaling regimes rather than asymptotic ones; this further weakens the unqualified infinite-hierarchy language. The location hierarchy (Eq. 6) and the FPT hierarchy for m≤3 may still stand, but the conclusion needs substantial qualification.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the extreme statistics of N independent random walks in a space-time random environment. It claims that the extreme location (maximum position) and the extreme first passage time (minimum time to reach a distant point) are governed by the Kardar-Parisi-Zhang (KPZ) equation, with means matching classical diffusion but variances receiving an environmental contribution that depends on the lowest random moment of the environment. The environmental variance is claimed to scale as in Eqs. (6) and (7), controlled by a parameter λ_ext, leading to an infinite hierarchy of universality classes indexed by m, where m−1 is the number of deterministic moments of the jump distribution. The paper derives these results through a replica-moment calculation of the tail probability, and supports them with numerical simulations for several distributions and for m=1,2,3,4.","tokens_in":21219,"tokens_out":4407,"duration_ms":43420,"significance":"If the results hold, they represent a significant extension of KPZ universality to higher moments of space-time disorder, and they give a concrete, measurable way to probe microscopic environment fluctuations through extreme-value statistics. The paper's strengths include the explicit identification of the generalized extreme diffusion coefficient D_ext^(m), the clean separation of sampling and environmental fluctuations, and a numerical comparison that uses independently measured environment statistics rather than fitted parameters. The numerical collapse in Fig. 2 for a wide class of distributions is encouraging. However, the claims as stated exceed what the paper's own analysis supports, particularly for the extreme first passage time and for the unqualified notion of an infinite hierarchy of universality classes.","major_comments":[{"comment":"The Numerical Results section explicitly states: 'For the extreme first passage time and m>3, my theoretical predictions break down.' Yet Eq. (7) is presented without any m≤3 restriction, the abstract claims the scalings of both the extreme location and the extreme first passage time depend on the moments, and the conclusion asserts an infinite hierarchy of universality classes. This is a load-bearing overstatement: the paper's own results establish the FPT hierarchy only for m=1,2,3, and the m>3 FPT claim is contradicted by the text. Moreover, for m>2 the environmental variance in Eq. (6) decays to zero as t→∞, so the higher 'universality classes' are transient scaling regimes rather than asymptotic ones. The revision should explicitly restrict Eq. (7) to m≤3, qualify Eq. (6) and the hierarchy language as transient for m>2, and adjust the abstract and conclusion accordingly.","section":"Numerical Results; Eq. (7); Abstract; Conclusion"},{"comment":"The derivation of the extreme first passage time relies on the non-backtracking approximation Pξ(τL≤t)≈Pξ(R(t)≥L), used without any quantitative justification. Since the paper itself reports that the FPT predictions break down in the bulk regime for m>3, and the same approximation underlies the m=1,2,3 FPT results, the validity of this approximation must be checked explicitly for the cases where the FPT hierarchy is claimed. Please provide a numerical or analytical verification of the non-backtracking approximation for m=2 and m=3, or state as a caveat that the FPT results inherit its limitations.","section":"Supplemental Material, Section VII"},{"comment":"The expansion of g(λ) is carried out explicitly only for m=1,2,3 and then the text states 'Thus, I extrapolate to higher m' to obtain Eq. (S19). The infinite-hierarchy claim therefore rests on an unproven pattern. This is acceptable as a conjecture, but the paper currently presents Eq. (S19) as a derivation. The revision should either prove the expansion for general m or explicitly mark Eqs. (6) and (7) as conjectural for m>3. The m=4 location numerics provide partial support, but they do not establish the general pattern.","section":"Supplemental Material, Section III A, Eq. (S19)"},{"comment":"The independence of SamN_t and EnvN_t is assumed and used to add variances in Eqs. (12)–(13). In the Supplemental Material the text states: 'Although I do not justify this here, my numerics indicate this is a reasonable assumption.' This is a structural assumption for all the main results. The numerical collapse in Fig. 2 is suggestive but does not constitute a proof. The revision should either prove this independence in the large-N, large-t (or large-L) limit or clearly identify it as an unproven assumption and discuss the possible consequences if it fails.","section":"Eqs. (12)–(13); Supplemental Material, Section VI"}],"minor_comments":[{"comment":"The phrase 'violates the first assumption, independence' is misleading: the random walks are independent given the environment; the environment induces correlations only in the sense that particles at the same site share the same jump distribution. Please rephrase for precision.","section":"Introduction"},{"comment":"The phrase 'The saturation of each curve is scaled by the diffusion coefficient, D' is unclear. Please specify the exact scaling used for the vertical axis in panels (b) and (e).","section":"Fig. 2 caption"},{"comment":"Equation (11) is stated as a condition for the simplified form of λ_ext, but the paper says it is only approximately satisfied for m>1 in the weak-noise limit. Please provide a quantitative bound or error estimate for this approximation, or state explicitly that the numerical agreement is used as justification.","section":"Main text after Eq. (8)"},{"comment":"The term 'super-universality class' is evocative but could be misunderstood; consider defining it as a hierarchy of universality classes indexed by the number of deterministic moments, with each class itself being a member of the KPZ universality class.","section":"Conclusion"}],"recommendation":"major_revision","confidential_remarks":"The manuscript overstates the 'infinite hierarchy' claim, particularly for the extreme first passage time where the paper's own numerics show a breakdown for m>3. However, the core derivation and the numerical evidence for the location hierarchy (including m=4) and for the FPT hierarchy up to m=3 appear sound. The issues are fixable by qualification and by making the extrapolation and independence assumptions explicit. I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline, in one breath: this is a real step beyond the known m=1 story, but the package needs to be sold with more care. The paper takes the established connection between extreme particle statistics in RWRE and the KPZ equation, generalizes it to environments where the m-th moment is the first random one, and writes down explicit variance formulas with a single prefactor lambda_ext. That is genuinely useful, and the numerics for m = 1 through 4 for the location, and m = 1 through 3 for the first-passage time, across several distinct environments, are convincing evidence that the pattern is real.\n\nThe main gap is the derivation of the general-m result. The expansion of g(lambda) is computed only for m = 1, 2, 3 and then extrapolated to all m. No general proof is supplied. The numerical checks make the extrapolation plausible, but it remains a conjecture, not a theorem, and the paper should say so plainly. The assumed independence of environmental and sampling fluctuations is also unproven; the numerics in Fig. 2 support it for the tested cases, so I would call that a moderate gap rather than a fatal one.\n\nThe more serious issue is the framing, and here the stress-test note is correct. The Numerical Results section explicitly says the first-passage-time predictions break down for m > 3, yet the abstract and conclusion claim an infinite hierarchy for both extremes. That overstates the support. The FPT hierarchy is at best established for m <= 3. Separately, for m > 2 the environmental variance decays to zero as t or L grows, so the higher-m regimes are transient scaling windows, not asymptotic universality classes. The paper does mention this in one paragraph, but the abstract and conclusion do not carry the qualification.\n\nFor the right reader — someone working on extreme statistics in disordered media, KPZ universality, or RWRE — this is worth attention. It is not a fully rigorous paper, but it is physically substantive, the formulas are checkable, and the numerical work is honest. It deserves peer review, though a serious referee should demand a clearer statement of what is proven versus conjectured, restrict the FPT claims to m <= 3, and reframe the higher-m behavior as crossover rather than asymptotic universality. I would send it out.","headline":"A solid extension of the m=1 KPZ extreme-diffusion results to higher moments, with explicit prefactors and honest numerics, but the 'infinite super-universal hierarchy' framing overstates both the FPT results (which fail for m>3 by the paper's own admission) and the asymptotic nature of the higher-m regimes.","tokens_in":21667,"tokens_out":3368,"would_cite":true,"duration_ms":37679,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["05.40.-a","05.40.Fb"],"model":"deepseek-v4-flash","headline":"The paper shows that the extremes of N diffusing particles in a space-time random environment are governed by the KPZ equation, with the variance scaling set by the lowest random moment of the environment.","keywords":["random walks in random environments","extreme value statistics","KPZ universality class","stochastic heat equation","first passage time","space-time random environment","super-universality class","outlier diffusion"],"falsifier":"Simulate an $m=2$ environment (zero drift, random per-site variance) and measure, over many environments, the covariance between $\\mathrm{Env}^N_t$ and $\\mathrm{Sam}^N_t$; if it is not zero, the variance addition in Eq. (12) is incomplete. Independently, check whether $\\mathrm{Var}_\\nu(\\mathrm{Env}^N_t)$ decays as $t^{-1/2}$ with the prefactor $(\\sqrt{2\\pi}/4)\\lambda_{\\mathrm{ext}}\\ln N$; a different exponent would falsify the identification of the noise strength of the effective KPZ equation.","tokens_in":2069,"feed_emoji":"🎲","tokens_out":3947,"duration_ms":164129,"temperature":0.7,"pith_summary":"This paper studies the most extreme particle in a cloud of N independent diffusing walkers when the medium itself is random in space and time. It argues that the extreme location and the extreme first-passage time are governed by the Kardar-Parisi-Zhang (KPZ) equation, and that the variance of these extremes splits into a universal sampling part and an environmental part. The environmental part is controlled by the lowest moment of the local jump distribution that still fluctuates: a random first moment acts as a random velocity field, a fixed first moment with a random second moment acts as a random diffusion coefficient, and fixing each further moment promotes the next one to set the scaling. Because each such regime is its own universality class, the family of models is called a super-universality class. The point matters because outlier statistics can expose microscopic fluctuations of the environment that ordinary diffusion theory ignores.","feed_headline":"One random environment moment sets outlier fluctuations","feed_subtitle":"Drift, diffusivity, skew: each lowest random moment creates its own universality class.","key_machinery":"The load-bearing object is the annealed tail probability $P_\\xi(R(t)\\ge x)$ and its convergence, after centering and rescaling, to the multiplicative stochastic heat equation (SHE), whose logarithm is the KPZ height. The paper identifies the noise strength of the SHE as $D_0=\\lambda_{\\mathrm{ext}}/((m!)^2(2D)^{(4m-1)/2})$, where $m-1$ is the number of deterministic moments of the single-site jump distribution and $\\lambda_{\\mathrm{ext}}=\\mathrm{Var}_\\nu(E_\\xi[Y^m])/(2E_\\nu[\\mathrm{Var}_\\xi(Y)])$ is the normalized variance of the lowest random moment. The argument hinges on expanding the collision term $g(\\lambda)$ of the tilted two-point motion; its leading term is $\\lambda^{2m}\\mathrm{Var}_\\nu(E_\\xi[Y^m])/(m!)^2$, which selects the lowest random moment and produces the anomalous scaling exponents in $t$ and $L$.","core_discovery":"On the paper's own terms, the central discovery is that the annealed tail probability of a single walker converges to the multiplicative stochastic heat equation, so the extremes of N walkers inherit KPZ fluctuations. The mean extreme values remain classical, $E[\\mathrm{Max}^N_t] \\approx \\sqrt{4Dt\\ln N}$ and $E[\\mathrm{Min}^N_L] \\approx L^2/(4D\\ln N)$. The variances are sums of a Gumbel sampling term, independent of the environment model, and an environmental term controlled by the lowest random moment $m$:  $\\mathrm{Var}_\\nu(\\mathrm{Env}^N_t) \\approx \\frac{\\sqrt{2\\pi}}{(m!)^2}\\lambda_{\\mathrm{ext}}(\\ln N)^{m-1}(Dt)^{(3-2m)/2}$ and $\\mathrm{Var}_\\nu(\\mathrm{Env}^N_L) \\approx \\frac{\\lambda_{\\mathrm{ext}}\\sqrt{\\pi}\\,2^{4m-9/2}}{(m!)^2D^2(\\ln N)^{(4m-9)/2}}L^{5-2m}$. Each $m$ defines a distinct scaling regime and hence a distinct universality class within the KPZ super-universality class; the numerical section reports collapse onto these formulas for a wide class of environments.","pith_inferences":["Beyond the paper's claims, the hierarchy suggests a spectroscopic use of extreme statistics: by measuring the outlier-variance exponent in $t$ or $L$, an experiment could infer which environmental moment is random without directly imaging the environment.","The paper itself reports that the first-passage formula breaks down for $m>3$ in the bulk regime because non-KPZ Gaussian fluctuations appear; completing the super-universality picture will require combining the present KPZ scaling with that bulk fluctuation theory.","A natural next test is to add temporal correlations to the environment; the SHE/KPZ limit would then carry a colored noise whose spectrum should appear as a frequency-dependent correction to the outlier variance.","Because the sampling variance is $m$-independent, subtracting the Gumbel term computed from two different $N$ values isolates the environmental term and gives a direct estimator of $\\lambda_{\\mathrm{ext}}$ without measuring the environment directly."],"forward_implications":["The mean extreme location and mean extreme first-passage time match classical diffusion, so the environment changes the spread of outliers rather than their typical position.","For $m=1$ the environmental variance grows with time or distance; for $m=2$ it decays for the location but still contributes a distinct $L^1$ term for the first-passage time; for $m>2$ it eventually vanishes, so the largest outliers become classically Gumbel-distributed at large scales.","The prefactor of every environmental variance term is set by $\\lambda_{\\mathrm{ext}}$, the normalized variance of the lowest random moment, so a single measured constant carries microscopic information about the environment.","Each integer $m$ defines a separate KPZ universality class; fixing one more moment moves the model one rung up the hierarchy, so the collection of all such environments is a single super-universality class.","If the assumed independence of environmental and sampling fluctuations holds, experimenters can separate the two contributions by comparing repeated runs in one environment with runs across many environments."],"supporting_citations":[{"why":"Defines the KPZ equation whose universality class is used to describe the extreme fluctuations.","marker":"[10]"},{"why":"Supplies the KPZ universality-class framework that organizes the scaling regimes.","marker":"[11]"},{"why":"Establishes the $m=1$ extreme-location fluctuations that the present hierarchy extends.","marker":"[12]"},{"why":"Provides the $m=1$ extreme first-passage-time method that this paper generalizes.","marker":"[13]"},{"why":"Introduces the extreme diffusion coefficient and the environmental/sampling variance decomposition used throughout.","marker":"[14]"},{"why":"Carries the moderate-deviation derivation for random walks in random environments that is extended to all $m$.","marker":"[15]"},{"why":"Connects diffusion in time-dependent random media to the KPZ equation.","marker":"[16]"},{"why":"Gives moderate-deviation scalings for the same model class, used to identify the KPZ scaling window.","marker":"[17]"},{"why":"Predicts Gaussian bulk fluctuations; the paper cites this to explain the $m>3$ first-passage breakdown.","marker":"[28]"}],"fun_headline_variants":["Each random moment sets a new universality class for outliers","KPZ extremes of outliers hinge on lowest random environment moment","Outlier diffusion: each random moment yields a distinct KPZ class","Random moment dictates extreme outliers' universality class","Super-universal: each random moment yields its own KPZ class for outliers"],"cache_read_input_tokens":23808,"weakest_assumption_plain":"The load-bearing premise is that the environment-dependent median Env and the within-environment sampling fluctuation Sam are independent, so their variances simply add in Eqs. (12)-(13); the paper says this independence is not proved and relies on numerical support, and if it fails the central variance formulas are incomplete.","fun_headline_variants_meta":{"raw":{"variants":["Each random moment sets a new universality class for outliers","KPZ extremes of outliers hinge on lowest random environment moment","Outlier diffusion: each random moment yields a distinct KPZ class","Random moment dictates extreme outliers' universality class","Super-universal: each random moment yields its own KPZ class for outliers"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001062,"raw_usage":{"total_tokens":4470,"prompt_tokens":980,"completion_tokens":3490,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":596,"completion_tokens_details":{"reasoning_tokens":3404}},"tokens_in":596,"tokens_out":3490,"duration_ms":24047,"temperature":1.0,"reasoning_tokens":3404,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T04:16:25.429707+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate an $m=2$ environment (zero drift, random per-site variance) and measure, over many environments, the covariance between $\\mathrm{Env}^N_t$ and $\\mathrm{Sam}^N_t$; if it is not zero, the variance addition in Eq. (12) is incomplete. Independently, check whether $\\mathrm{Var}_\\nu(\\mathrm{Env}^N_t)$ decays as $t^{-1/2}$ with the prefactor $(\\sqrt{2\\pi}/4)\\lambda_{\\mathrm{ext}}\\ln N$; a different exponent would falsify the identification of the noise strength of the effective KPZ equation.","supporting_citations":[{"cited_title":"Dynamic Scaling of Growing Interfaces","cited_arxiv_id":null,"evidence_quote":"Defines the KPZ equation whose universality class is used to describe the extreme fluctuations."},{"cited_title":"The Kardar–Parisi–Zhang Equation and Universality Class","cited_arxiv_id":null,"evidence_quote":"Supplies the KPZ universality-class framework that organizes the scaling regimes."},{"cited_title":"Hass, Aileen N","cited_arxiv_id":null,"evidence_quote":"Establishes the $m=1$ extreme-location fluctuations that the present hierarchy extends."},{"cited_title":"b,e) show the numerically measured environ- mental variance asymptotes to my theoretical predictions for a wide range of distributions and m","cited_arxiv_id":null,"evidence_quote":"Provides the $m=1$ extreme first-passage-time method that this paper generalizes."},{"cited_title":"Hass, Hindy Drillick, Ivan Corwin, and Eric I","cited_arxiv_id":null,"evidence_quote":"Introduces the extreme diffusion coefficient and the environmental/sampling variance decomposition used throughout."},{"cited_title":"Hass, Ivan Corwin, and Eric I","cited_arxiv_id":null,"evidence_quote":"Carries the moderate-deviation derivation for random walks in random environments that is extended to all $m$."},{"cited_title":"Diffusion in Time-Dependent Random Media and the Kardar-Parisi- Zhang Equation","cited_arxiv_id":null,"evidence_quote":"Connects diffusion in time-dependent random media to the KPZ equation."},{"cited_title":"Universal KPZ Fluctuations for Moderate Devia- tions of Random Walks in Random Environments, March 2025","cited_arxiv_id":null,"evidence_quote":"Gives moderate-deviation scalings for the same model class, used to identify the KPZ scaling window."},{"cited_title":"Conclusion","cited_arxiv_id":null,"evidence_quote":"Predicts Gaussian bulk fluctuations; the paper cites this to explain the $m>3$ first-passage breakdown."}],"review_version":1}