{"id":"a4b809f4-f155-4dc3-a1dd-6bd6efacaabf","arxiv_id":"2412.19359","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The transit-length distribution for a binary Markovian mixture is the telegraph-process occupation-time distribution; the paper adds a stable high-mixing asymptotic form that converges to atomic mix and improves porous-slab transmission estimates.","lead":"This paper derives a formula for how far a particle travels inside one material when it crosses a random two-material mixture, using the same equations as the telegraph process. It also provides a fast, numerically stable approximation for well-mixed materials and shows it outperforms the simple atomic-mix average on porous slab examples.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Section 3's light-transmission application has a missing cosine factor: Eq. (55)/(56) should be g(μ)e^{-Σ_t ℓ}, not g(μ)e^{-Σ_tℓ/μ}; Eq. (58) is consistent with the corrected form, so the exact and asymptotic results in §3 are internally inconsistent for μ ≠ 1.","rationale":"The stress-test pass confirms the main derivation: the Neumann-sum/Bessel reduction leading to Eq. (24) is algebraically coherent, the transformation ℓ→u in Section 2.2 carries the correct Jacobian, and the reported Monte Carlo agreement is meaningful evidence for the implementation. The reader's identified assumption (Markovian exponential chords) is a declared modeling premise, not an internal defect. However, Section 3 contains a concrete internal inconsistency that is more load-bearing for the paper's stated applications. From Eq. (53), the correct attenuation along a ray is exp[-Σ_tℓ], not exp[-Σ_tℓ/μ]; the printed Eq. (55) makes the final equality wrong by a factor μ, and Eq. (56) inherits it. The asymptotics in Eq. (58) use the correct exp[-Σ_tℓ], as shown by the leading factor e^{-ωΣ_tx/μ}, so the exact and approximate models are not approximating the same integral for μ≠1. The pure-absorber limit establishes this independently of the approximate forms. The beam-source parameter studies are unaffected because μ=1, but the isotropic-flux results and any angularly integrated comparisons are suspect. Since the core distribution is sound and the defect is localized to Section 3, the appropriate disposition is conditional acceptance: require correction of Eq. (55)/(56) (or explicit clarification that ℓ is redefined as projected length, together with the corresponding Jacobian in f_A) and a rerun of the affected numerical results. I do not see grounds for rejection of the transit-length derivation itself.","tokens_in":19944,"tokens_out":22055,"duration_ms":195915,"concrete_test":"Recompute the isotropic-flux transmission probability in Section 3.3.1, replacing the kernel in Eq. (56) by g(μ)e^{-Σ_tℓ} (and correcting the exact evaluation in the repository accordingly), and compare the resulting transmission probability with the published Fig. 5 and with the asymptotic Eq. (62). Also run the pure-absorber check ω=1: the corrected formula must return g(μ)e^{-Σ_t X/μ}, while Eq. (56) as printed returns g(μ)e^{-Σ_t X/μ^2}. If the corrected Fig. 5 differs materially from the published curve, Section 3 and the associated repository scripts need revision; if it does not differ, Eq. (55) is a typographical error and the conclusions can stand after relabeling.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The load-bearing soft spot is in Section 3, not in the Section 2 derivation. Starting from Eq. (53), μ∂ψ/∂x + Σ_t(x)ψ = 0, a ray with direction μ has path increment ds = dx/μ. The attenuation is exp[-∫Σ_t ds] = exp[-Σ_t ℓ], where ℓ is the along-ray length in material. The middle expression in Eq. (55), exp[-(1/μ)∫_0^x Σ_t(x')dx'], equals exp[-Σ_t ℓ]; the final equality in Eq. (55), g(μ)e^{-Σ_tℓ/μ}, has an extra factor 1/μ. The same factor enters the exact ensemble average in Eq. (56). The limiting case ω=1 (pure absorber) makes this transparent: f_A(ℓ,x/μ)=δ(ℓ-x/μ), and Eq. (56) gives g(μ)e^{-Σ_t x/μ^2} instead of the standard g(μ)e^{-Σ_t x/μ}. Moreover, the asymptotic expansion Eq. (58) has leading term e^{-ωΣ_t x/μ}, which is the expansion of e^{-Σ_tℓ} about ℓ(0)=ωx/μ, not of e^{-Σ_tℓ/μ}. Thus the approximate formula in Eqs. (58)-(60) is inconsistent with the exact formula in Eq. (56) it is supposed to approximate whenever μ ≠ 1. For the beam-source cases μ=1 the error cancels, so Figs. 4, 6-9 and the charged-particle sections are not necessarily affected; but the isotropic-flux exact results in Fig. 5 and any μ-averaged transmission comparisons are suspect. This does not undermine Eq. (24)/(40) or the Monte Carlo verification; it undermines the claimed broad applicability of the light-transmission application.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper derives an exact distribution for the transit length ℓ that a particle spends in material A while traveling a fixed total distance s through a binary Markovian random mixture. The authors map the problem onto a telegraph process in which the net distance X_A = 2ℓ − s plays the role of position and the total path length s plays the role of time, solve the resulting two-state first-order PDE system by the method of characteristics, and sum the Neumann series in the number of zone crossings to obtain the Bessel-function densities in Eqs. (22a)–(24), including Dirac masses for particles that never leave their starting zone. For strongly mixed media they derive an asymptotically Gaussian density in the variable u = √ℓ − √(k(s−ℓ)) (Eqs. (31)–(40)), with renormalization fixed by a unit-integral condition, and they show analytically that the asymptotic forms reduce to a Dirac delta at the atomic-mix transit length as α → ∞. Numerical support includes a 10^9-sample Monte Carlo simulation of the Markov process (agreement within 1σ, Fig. 2) and an L1 error study showing α^{−1} convergence of the asymptotic form (Fig. 3). The distribution is then applied to two porous-media problems: light transmission through a purely absorbing slab, for which an exact ensemble-averaged angular flux (Eq. (56)) and an analytic approximate form (Eqs. (58)–(62)) are derived; and charged-particle stopping, for which transmission probabilities, energy spectra, and angular fluxes are computed.","tokens_in":20237,"tokens_out":29554,"duration_ms":234185,"significance":"The Section 2 derivation is internally consistent: the characteristic integrations, the Bessel-function resummation, the normalization constants, and the atomic-mix limit all check out, and the Monte Carlo and L1-norm evidence supports the claims. The paper ships reproducible implementations (C++, Python, MATLAB) and all scripts, and it states its own limitations explicitly, notably in §4.1.1, where the asymptotic transmission model is honestly restricted to the atomic-mix-transmitting regime and its failure for thick slabs is demonstrated. These are genuine strengths. However, the light-transmission application contains a real error for μ ≠ 1: the printed ensemble average uses e^{−Σ_tℓ/μ} where the physics requires e^{−Σ_tℓ}, so the isotropic-flux exact results in Fig. 5 and any μ-averaged statements in Section 3 are invalid as printed, while the beam-source results (μ = 1) are unaffected. The central printed formula Eq. (24) also carries a support-sign typo in its Heaviside factor. Both issues are local and fixable, and they do not touch the Section 2 derivation or the charged-particle application.","major_comments":[{"comment":"The Heaviside factor Θ(ℓ(ℓ−s)) in the continuous parts of Eqs. (22a), (22b), and (24) has the wrong sign. With the paper's definition of Θ (unity for non-negative argument), Θ(ℓ(ℓ−s)) vanishes for 0 < ℓ < s, so the printed densities are zero on their claimed support. From Eqs. (17) and (20), the factor before the change of variables is Θ((s−x)(s+x)), which becomes Θ(4ℓ(s−ℓ)) = Θ(ℓ(s−ℓ)) under x = 2ℓ − s; the correct support is 0 ≤ ℓ ≤ s. Because the Monte Carlo verification in Fig. 2 and all subsequent numerical results evidently used the correct support, this is presumably a typographical slip, but it must be corrected because a reader implementing Eq. (24) as printed would obtain a density that vanishes almost everywhere on (0, s).","section":"§2.1, Eqs. (22a), (22b), (24)"},{"comment":"For a ray with direction cosine μ, the path increment along the ray is ds = dx/μ, and the projected length of the A-segments is μℓ, where ℓ is the along-ray transit length in A; hence ∫_0^x Σ_t(x′)dx′ = Σ_t μℓ and the middle expression exp[−(1/μ)∫Σ_t dx′] equals e^{−Σ_tℓ}, not e^{−Σ_tℓ/μ}. The second equality in Eq. (55) is therefore inconsistent with the first for μ ≠ 1, and the ensemble average in Eq. (56) inherits the error. The limiting case ω → 1 makes this crisp: f_A(ℓ, x/μ) = δ(ℓ − x/μ) and Eq. (56) gives g(μ)e^{−Σ_t x/μ²} instead of the standard g(μ)e^{−Σ_t x/μ}. Independently, the asymptotic expansion in Eq. (58) has leading factor e^{−ωΣ_t x/μ}, which is the expansion of e^{−Σ_tℓ} about ℓ(0) = ωx/μ, so the approximate solution is consistent with the corrected attenuation rather than with Eq. (56). The exact and approximate results in Section 3 are thus internally inconsistent whenever μ ≠ 1. The beam-source results (Figs. 4 and 6–9) are unaffected because μ = 1, but the isotropic-flux exact curve in Fig. 5 and any μ-averaged transmission comparisons must be recomputed after replacing e^{−Σ_tℓ/μ} with e^{−Σ_tℓ} in Eqs. (55)–(56).","section":"§3.1, Eqs. (55)–(56)"},{"comment":"The asymptotic derivation states that 'we require that α is large, and since k is on the order of or greater than 1' (text following Eq. (26)), yet the applications use k = 0.2 (§3.3.1), sweep k from 0 to 10 (§3.3.2), and use k = 0.25 (§4.3). The negative-density analysis of Eqs. (34) and (38) covers the regimes k > 3 and k < 1/3, and the truncation bounds in Eqs. (42) are designed for the k < 1 cases, but the paper never states the regime of validity of the asymptotic model as a function of k, nor analyzes the error introduced by the truncation when k < 1. As printed, the validity of Eqs. (40)–(43) at k = 0.2 and k = 0.25 rests entirely on the numerical comparisons in Figs. 4–9. I ask the authors to make the domain of validity explicit (including whether Eq. (42) fully removes the negative-density defect) or to qualify the accuracy claims for k substantially below one.","section":"§2.2 vs. §§3.3 and 4.3"}],"minor_comments":[{"comment":"The stated negativity onset u < −2√(ks)/(k+1) is not the zero of the prefactor 1 + (1/2)√((1+k)/(ks))u; the zero occurs at u = −2√(ks/(1+k)). The analogous bound after Eq. (38) should likewise be u > 2√(ks/(1+k)). The threshold conditions k > 3 and k < 1/3 are correct, and only the locations are misstated.","section":"§2.2, text after Eq. (34)"},{"comment":"The domain is given as '0 ≤ x ≤ L' although the slab width is denoted X throughout Section 3; use one symbol consistently.","section":"§3.1, Eq. (53)"},{"comment":"Typos: 'Isotopic flux' appears in the captions of Figs. 5, 14, and 15 and in §3.3.1 and should be 'isotropic flux'; 'chracteristic' in the Fig. 1 caption; 'The transport of light though this problem' in §3.1; 'a vector of of evaluation points' in §2.4.4; 'This asympotic solution' in §4.1.1; 'Theortical' in reference [4]; 'Generized' in reference [5].","section":"Throughout"},{"comment":"The description 'a grid of 100 even cosine µ intervals' is unclear; please state the integration rule (for example, uniform in µ) and the quadrature used to produce the isotropic-flux results in Fig. 5.","section":"§3.3.1"},{"comment":"The sentence 'The L-1 norm is computed shown as a function of α' contains a redundant word, and the text discusses only the asymptotic error curve in Fig. 3 although the caption also distinguishes the atomic-mix error.","section":"§2.5.2"}],"recommendation":"major_revision","confidential_remarks":"The Section 2 result is sound and well verified, and the paper is a good fit for a mathematical-physics audience; the problems are localized and fixable. The μ-factor error in Section 3 is real — I verified both the pure-absorber limiting check and the consistency of Eq. (58) with the corrected form — so the isotropic-flux results must be redone, but the beam-source claims stand. The Heaviside support typo in Eq. (24) and the unqualified k < 1 usage are also easily addressed. I therefore recommend major revision rather than rejection. On novelty: the transit-length density itself is closely related to Ratanov's occupation-time law for the telegraph process, and the authors already cite this; the incremental novelty lies in the transport framing, the stable asymptotics, and the applications, which I consider sufficient for the journal if the revisions are made."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Core take: the transit-length density itself is a solid result, and the numerically stable high-mixing asymptotics are genuinely useful. The Section 2 derivation is clean, the 1e9-sample Monte Carlo check is convincing, and shipping code in three languages helps reproducibility. Also good: the authors cite Ratanov and acknowledge the telegraph-process overlap, so the bare density is not oversold; the new value is in the asymptotics and the porous-media applications.\n\nThe soft spot is in Section 3. The transport equation μ∂ψ/∂x + Σ_tψ = 0 attenuates along a ray as exp(-Σ_t ℓ), where ℓ is the actual distance spent in the absorber. Eq. (55)–(56) instead write exp(-Σ_t ℓ/μ). That extra 1/μ is wrong. The asymptotic expansion in Eq. (58) has leading factor e^{-ωΣ_t x/μ}, which is the expansion of exp(-Σ_t ℓ) about ℓ(0)=ωx/μ, not of exp(-Σ_t ℓ/μ). So the exact and approximate forms in Section 3 are internally inconsistent for μ ≠ 1. For beam sources (μ=1) the error cancels, so Figs. 4, 6–9 and the charged-particle sections are likely fine. But the isotropic-flux results in Fig. 5 and any μ-averaged comparisons need to be redone. This is fixable, but it is a real load-bearing error for that application.\n\nMinor notes: the asymptotic form's first-order Taylor linearization of the fourth-root term is acknowledged to be limited to near-peak regions, and the charged-particle section honestly states when the approximation fails. The L1 error study shows α^{-1} convergence, which matches the claim. The Markovian exponential-chord assumption is a modeling premise, not a consequence; that's fine given the title and scope.\n\nWho this is for: people working on stochastic-media transport, particularly in radiation transport and porous media. It deserves a serious referee, but the referee should require the Section 3 cosine factor to be fixed and affected figures regenerated. My own verdict would be major revision, not acceptance as is.","headline":"Useful exact distribution and high-mixing asymptotics, but Section 3 has a wrong 1/μ attenuation factor that undercuts the non-beam light-transmission results.","tokens_in":20840,"tokens_out":8046,"would_cite":true,"duration_ms":59523,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["82C70","60K40"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper derives an exact transit-length distribution for particles in binary Markovian mixtures, with a stable asymptotic form that beats atomic-mix in porous media.","keywords":["transit-length distribution","binary Markovian mixtures","telegraph process","stochastic media","porous media","atomic-mix limit","modified Bessel functions","particle transport"],"falsifier":"Simulate or measure the transit-length density in a binary mixture where chord lengths are drawn from distributions with the same means but nonzero correlations or nonexponential tails; if the measured density departs from Eq. (24) beyond Monte Carlo or experimental uncertainty, the Markovian premise is refuted. A cheaper check: compare the predicted L1 convergence rate of the asymptotic form (roughly $\\alpha^{-1}$) against a high-resolution Monte Carlo for $\\alpha$ from 10 to 200 cm$^{-1}$ in the paper's slab test case.","tokens_in":19648,"feed_emoji":"⚛️","tokens_out":7482,"duration_ms":69773,"temperature":0.7,"pith_summary":"This paper derives an exact probability density for the transit length: the distance a particle traveling a straight-line path of total length $s$ spends inside one component of a binary random mixture. The model assumes each material zone has an exponentially distributed chord length, with transition rates $\\alpha$ and $\\beta$, and it exploits an equivalence between transport in such mixtures and the telegraph random process. The central result, Eq. (24), is a closed form involving modified Bessel functions $I_0$ and $I_1$, valid for arbitrary starting material and total distance. Because the Bessel form overflows for highly mixed media, the paper also derives a numerically stable asymptotic form, Eq. (40), and shows it converges to the atomic-mix limit. A sympathetic reader would care because this turns a previously numerical or approximate task -- computing path-length-dependent transmission and spectra in porous media -- into an analytic or high-accuracy computation.","feed_headline":"Particle travel in random mixtures now has an exact density","feed_subtitle":"A stable asymptotic form converges to atomic-mix and beats it for highly mixed porous slabs, from light to electron stopping.","key_machinery":"The machinery is the mapping of binary Markovian transport onto the telegraph process: total distance $s$ plays the role of time, net distance in material A minus material B plays the role of position, unit speed is used, and the zone-crossing rates $\\alpha,\\beta$ play the role of switching rates. The derivation solves a Neumann expansion of first-order PDEs for $p_A^m(x,s)$ and $p_B^m(x,s)$, transforms them with an exponential ansatz, and sums the series to closed form via modified Bessel functions. The asymptotic analysis uses the large-argument form $I_n(z)\\sim e^z/\\sqrt{2\\pi z}$ together with the substitution $u=\\sqrt{\\ell}-\\sqrt{k(s-\\ell)}$ to expose a Gaussian factor $e^{-\\alpha u^2}$; a first-order Taylor expansion of the remaining fourth-root factor yields the analytically tractable skewed Gaussian of Eq. (40), with renormalization constants computed by enforcing unit probability.","core_discovery":"The paper's central claim is that for a binary Markovian mixture with exponential chord-length rates $\\alpha=1/\\Lambda_A$ and $\\beta=1/\\Lambda_B$, the density of the distance $\\ell$ spent in material A after total travel $s$, starting from material A or B, is given by Eqs. (22a)-(22b), with the mixed-start version in Eq. (24). The density is a superposition of Dirac deltas for particles that never leave their starting material, plus continuous terms built from modified Bessel functions $I_0$ and $I_1$; the $I_0$ terms count odd numbers of zone crossings and the $I_1$ terms count positive even numbers. The paper further claims that in the highly mixed limit the density approaches a skewed Gaussian (Eq. (40)) in the variable $u=\\sqrt{\\ell}-\\sqrt{k(s-\\ell)}$, and that this asymptotic form converges to the Dirac delta at the atomic-mix transit length $\\ell = ks/(1+k)$ as $\\alpha\\to\\infty$. Monte Carlo sampling of the exponential telegraph process confirms the exact density, and the applications show the asymptotic model is more accurate than atomic mix for highly mixed, not-too-thick porous slabs.","pith_inferences":["Beyond the paper, the same density could be used to construct reaction-rate path-length estimators for Monte Carlo particle transport that average over material randomness inside a single flight, potentially lowering variance; the paper mentions this as future work, not a demonstrated result.","Beyond the paper, if the Markovian premise is replaced by correlated or nonexponential chord lengths, the closed Bessel form will not hold, but the general strategy of solving a hierarchy of crossing-count equations might extend to renewal-type mixtures, yielding a generalized transit-length density.","Beyond the paper, the asymptotic skewed-Gaussian form suggests that in highly mixed media the transit-length statistics depend on the single combination $\\alpha$ times an effective quadratic form, so error estimates for homogenization could be expressed in terms of $\\alpha s$ and $k$ alone.","Beyond the paper, one could test the model against transmission measurements in fabricated random slabs with engineered chord-length statistics; agreement for exponential statistics and systematic deviation for non-exponential statistics would isolate the role of the Markovian assumption."],"forward_implications":["Ensemble-averaged angular flux in a purely absorbing porous slab can be written exactly as an integral of the transit-length density (Eq. (56)) and, under the asymptotic model, reduced to standard functions (Eq. (62)).","For highly mixed, not-too-optically-thick slabs, the analytic asymptotic transmission probability is more accurate than the atomic-mix result, reaching sub-0.1% error for beam sources at modest $\\alpha$ while atomic mix requires much larger $\\alpha$.","Charged-particle stopping in porous media can be treated with the same distribution: transmission probability, energy spectrum, and angular flux all follow from integrating the transit-length density against the stopping-power relation, with the caveat that the analytic asymptotic transmission form is reliable only when particles transmit in the atomic-mix limit.","The exact density is verified against a $10^9$-sample Monte Carlo simulation of the exponential telegraph process, so the formulas can serve as reference solutions for testing approximations in stochastic media transport.","Because the asymptotic density converges to the atomic-mix limit, it provides a controlled bridge between full stochastic-media models and simple volumetric homogenization."],"supporting_citations":[{"why":"It supplies the standard binary-mixture transport closure with exponential chord lengths that this work builds on.","marker":"[1]"},{"why":"It establishes the link between the telegrapher's equation and a switching random process that the paper exploits.","marker":"[7]"},{"why":"It derives an occupation-time density for a telegraph process that the paper adapts to transit lengths.","marker":"[8]"},{"why":"It applies the transit-length distribution as a weighting function for electron transport, the source of the authors' earlier results.","marker":"[9]"},{"why":"It treats energy-loss straggling in the same binary Markovian setting, contextualizing the stopping-power application.","marker":"[10]"}],"fun_headline_variants":["Exact transit-length density for particles in random mixtures","Beyond atomic mix: exact transit lengths in random media","Exact transit distances for light and electron stopping in random slabs","Monte Carlo validates exact transit-length density in random media"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the distance a particle travels inside either material before crossing to the other is exponentially distributed with fixed rates $\\alpha$ and $\\beta$; if real mixtures have correlated or otherwise nonexponential chord-length statistics, the telegraph-process equivalence, the closed-form density, and both applications no longer follow.","fun_headline_variants_meta":{"raw":{"variants":["Exact transit-length density for particles in random mixtures","Beyond atomic mix: exact transit lengths in random media","Exact transit distances for light and electron stopping in random slabs","Monte Carlo validates exact transit-length density in random media"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001429,"raw_usage":{"total_tokens":5768,"prompt_tokens":949,"completion_tokens":4819,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":565,"completion_tokens_details":{"reasoning_tokens":4762}},"tokens_in":565,"tokens_out":4819,"duration_ms":34043,"temperature":1.0,"reasoning_tokens":4762,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T00:40:47.888174+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate or measure the transit-length density in a binary mixture where chord lengths are drawn from distributions with the same means but nonzero correlations or nonexponential tails; if the measured density departs from Eq. (24) beyond Monte Carlo or experimental uncertainty, the Markovian premise is refuted. A cheaper check: compare the predicted L1 convergence rate of the asymptotic form (roughly $\\alpha^{-1}$) against a high-resolution Monte Carlo for $\\alpha$ from 10 to 200 cm$^{-1}$ in the paper's slab test case.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies the standard binary-mixture transport closure with exponential chord lengths that this work builds on."},{"cited_title":"Kac , A Stochastic Model Related to the Telegrapher's Equation, Rocky Mountain J","cited_arxiv_id":null,"evidence_quote":"It establishes the link between the telegrapher's equation and a switching random process that the paper exploits."},{"cited_title":"Ratanov , A Jump Telegraph Model for Option Pricing, Quantitative Finance, 7, 5, 575 (2007)","cited_arxiv_id":null,"evidence_quote":"It derives an occupation-time density for a telegraph process that the paper adapts to transit lengths."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It applies the transit-length distribution as a weighting function for electron transport, the source of the authors' earlier results."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It treats energy-loss straggling in the same binary Markovian setting, contextualizing the stopping-power application."}],"review_version":1}