{"id":"9835caa7-f80f-41e0-ba9f-6bf7d2ddac96","arxiv_id":"2411.10360","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A new radiant field theory derives the transmission eigenvalue distribution for waves through disordered waveguides beyond the diffusive regime.","lead":"The authors introduce a transport equation for a matrix-valued radiance that predicts how much light passes through a disordered material, without assuming isotropic scattering. It captures the quasiballistic regime, absorption, and partial control of input channels, which earlier theories missed.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The single-replica saddle-point approximation is asserted without an error estimate; its failure for smaller channel numbers or near localization would invalidate Eq. (13).","rationale":"The reader's weakest assumption identifies the same single-replica saddle-point approximation as the most delicate step. I agree that this is the load-bearing concern. My partial disagreement is that the reader emphasizes the localization boundary (L ≈ ξ), while the more concrete and testable risk is the finite channel number N: the approximation is a mean-field (saddle-point) treatment that should become exact as N → ∞, but the Letter only tests N ≈ 100. The proposed test directly probes the regime where corrections of order 1/N are expected to appear. The paper's strong points — reproducible code, the exact DMPK reduction in the diffusive limit, and agreement with wave-equation numerics for N ≈ 100 — justify a conditional acceptance pending this check. An unchanged CONDITIONAL verdict is therefore appropriate, rather than ACCEPT or REJECT.","tokens_in":9950,"tokens_out":7485,"duration_ms":78098,"concrete_test":"Run the RFT solver (Ref. [41]) and the recursive Green's function code (Ref. [42]) for a waveguide with W/λ ≈ 5 (≈10 propagating modes) at L/ℓ = 5 and L/ℓ = 0.2, and compare the low-transmission tail (T < 0.01) of ρ(T). If the RFT deviates from the wave-equation numerics by more than the statistical error bars in this finite-N setting, the replica-decoupling approximation is not uniformly valid; if agreement persists, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim rests on the derivation of Eq. (13), and the most load-bearing step is the assertion that the replicated partition function factorizes, <Z^R> ≈ (bar{Z})^R, by neglecting inter-replica coupling (text after Eq. (11)). This is a saddle-point plus replica-decoupling approximation, stated to be valid for L << ξ but with no quantitative error bound. In a quasi-1D waveguide, the number of propagating channels N is finite (N ≈ 100 in the figures); corrections to the replica-symmetric saddle point are generally of order 1/N or 1/(kℓ). Such corrections may be hidden in the plotted agreement for N≈100 but could become visible for smaller N or in the extreme tails of ρ(T), where the generating function F(γ) is evaluated at γ = 1/T + i0+ and non-perturbative contributions can matter. The Letter's numerical validations all use W/λ = 50.5, so they do not probe the finite-N sensitivity of the replica-decoupling assumption. Without an independent check of this approximation, the claim that Eq. (13) is universally accurate in the quasiballistic regime is not fully secured.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents a field-theoretic framework, 'radiant field theory' (RFT), for the distribution of transmission eigenvalues of coherent waves in a disordered waveguide. Starting from a microscopic wave equation with a white-noise random potential, the authors use a replica trick, a Hubbard-Stratonovich transformation, a saddle-point approximation, and a Wigner transform to derive a matrix transport equation for a matrix-valued radiance g(Ω,r) [Eq. (13)], with boundary conditions [Eq. (15)] and a generating function [Eq. (17)]. The theory is validated by comparing its predictions with numerical solutions of the wave equation in three settings: varying optical thickness L/ℓ, uniform absorption, and partial channel control. The authors also show that a direction-averaged limit of Eq. (13) reproduces the DMPK equation in the diffusive regime.","tokens_in":10185,"tokens_out":7136,"duration_ms":80918,"significance":"If the central equation (13) is valid, RFT is a substantial generalization of DMPK theory: it removes the isotropy hypothesis, extends predictions into the quasiballistic regime, and incorporates absorption and incomplete channel control in a parameter-free way. The paper is strengthened by concrete numerical evidence from the microscopic wave equation, by the exact diffusive-limit benchmark against DMPK, and by the release of two numerical codes. The main unresolved question is the rigor of the single-replica saddle-point step, which is asserted but not quantitatively controlled in the Letter.","major_comments":[{"comment":"The factorization ⟨Z^R⟩ ≃ (Z̄)^R, obtained by neglecting inter-replica coupling, is the step that reduces the 2R×2R saddle-point problem to the one-replica matrix transport equation (13). The Letter states that this is valid in the nonlocalized regime L ≪ ξ but gives no error estimate. Since Eq. (13) is the central object from which all predictions follow, this is a load-bearing gap. I request a quantitative justification (for instance, an estimate in powers of 1/(kℓ) or 1/N) or, failing that, an explicit numerical test of the factorization's accuracy.","section":"Derivation, after Eq. (11)"},{"comment":"All numerical validations use W/λ = 50.5 and L/W = 1, so only a single channel count N ≈ 100 is probed. Corrections to the replica-symmetric saddle point are generically of order 1/N, and the plotted agreement could in principle mask a finite-N sensitivity. A test at a substantially smaller width (e.g., W/λ ≈ 10) or at a different aspect ratio would materially strengthen the claim that Eq. (13) is accurate in the quasiballistic regime and is not tied to the specific large-N geometry used here.","section":"Figs. 1–3"},{"comment":"The derivation of the Wigner-transport equation (13), the boundary conditions (15), and the generating function (17) is delegated to the companion paper. While this division is acceptable for a Letter, the text should state more precisely the approximations entering the on-shell Wigner reduction and their expected small parameters. In particular, the 'semiclassical' approximation is not accompanied by a statement of its error in the quasiballistic regime, so the reader cannot distinguish a controlled expansion from an uncontrolled one.","section":"Eqs. (12)–(17) and companion paper [38]"}],"minor_comments":[{"comment":"The sentence 'with N a γ-independent arbitrary prefactor' is confusing because N is also used for the number of channels; please denote the prefactor by, e.g., calligraphic N or explain the notation explicitly.","section":"Eq. (4) and surrounding text"},{"comment":"The 'dashed integral' notation is used without a definition in the main text; please define it as the Cauchy principal value consistently, or refer explicitly to the companion paper at the point of first use.","section":"Eq. (12)"},{"comment":"There is a formatting inconsistency: 'W/λ = 50 .5' appears with a stray space; similar spacing issues occur in other figure captions (e.g., '10−1 100 101').","section":"Fig. 3 caption"},{"comment":"The claim that Eq. (19) 'exactly reduces to the DMPK solution' is stated without derivation or reference to a specific equation in the companion paper; a short derivation or precise pointer would make the benchmark check easier to verify.","section":"Text after Eq. (19)"}],"recommendation":"major_revision","confidential_remarks":"The companion paper [38] is cited as accepted in Phys. Rev. Res., and the numerical codes are publicly available. My main concern is that the central replica-decoupling step is asserted rather than controlled, and that the numerical tests do not vary the channel number. If the companion paper contains a rigorous derivation and finite-N checks, then the requested changes are mostly presentational; otherwise the Letter's central claim is not fully secured."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: the Letter contains a genuinely new object — the matrix transport equation (13) — that extends Eilenberger/Usadel machinery to transmission eigenvalue statistics without the isotropy hypothesis, and it ships with enough numerical evidence to be taken seriously. It deserves a serious referee, not a desk rejection.\n\nWhat is new: rather than assuming uniform angular scattering per slice, the theory keeps the full direction-resolved radiance, so it remains applicable in the quasiballistic regime, and it incorporates absorption and partial channel control in a natural way. The diffusive limit reduces to DMPK/Usadel, which is a good external sanity check. The Letter's own numerics — recursive Green's function simulations of the wave equation for L/ell from 0.05 to 5, with and without absorption and with reduced apertures — match the theory remarkably well, and the deviating DMPK/FRM dashed lines make the improvement visible. The code and data are on Zenodo, so the numerical part is reproducible. That is substantial credit.\n\nThe soft spot is the derivation. Eq. (13) is the central claim, but the derivation in this Letter is a sketch: replica averaging, Hubbard-Stratonovich, saddle point, Wigner transform. The load-bearing approximation is the replica decoupling <Z^R> ≈ Zbar^R and the single-replica saddle point, asserted to be valid for L << xi without an error estimate. All tests use W/lambda = 50.5, so N ~ 100; corrections of order 1/N or 1/(k ell) could be hidden, and the extreme tails of rho(T) are where non-perturbative effects could surface. The companion paper apparently carries the full derivation, and its acceptance in PRR gives some assurance, but from the Letter alone the central step is not independently checkable. I would not call this fatal — in the large-N nonlocalized regime the approximation is plausible and the numerics support it — but it is a real gap in the Letter itself.\n\nThe citation pattern looks fair: DMPK, sigma models, Nazarov, Brouwer, FRM, and the companion paper are all there, and the self-citation to the companion is legitimate since that is where the derivation lives.\n\nWho is this for: people working on wavefront shaping, mesoscopic transport, or random-matrix approaches to transmission. The framework gives them a new tool with code. I would send it to peer review. A referee should push on the replica-decoupling step and on finite-N checks, but this is exactly the kind of work that merits referee time.","headline":"Genuinely new, reproducible matrix transport theory that nails transmission eigenvalue statistics beyond the diffusive regime; the replica-decoupling step is the open sore, but the Letter deserves serious peer review.","tokens_in":10679,"tokens_out":2513,"would_cite":true,"duration_ms":25023,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["82B44","81Q20","78A45"],"pacs":[],"model":"deepseek-v4-flash","headline":"A matrix-valued radiance obeying a nonlinear transport equation determines the transmission eigenvalue distribution of a disordered waveguide across diffusive and quasiballistic regimes, with absorption and partial channel control included.","keywords":["transmission eigenvalues","disordered waveguides","matrix transport equation","matrix-valued radiance","quasiballistic transport","absorption","incomplete channel control","replica method"],"falsifier":"Solve Eq. (13) for a waveguide with optical thickness $L/\\ell$ fixed and gradually push the ratio $L/\\xi$ toward and beyond one by increasing disorder strength or shrinking the width, then compare with recursive Green's function simulations; if the RFT prediction departs from the microscopic distributions as localization is approached, the single-replica saddle-point approximation is the point of failure.","tokens_in":9746,"feed_emoji":"🌊","tokens_out":6670,"duration_ms":62434,"temperature":0.7,"pith_summary":"The paper claims that the full distribution of transmission eigenvalues for coherent waves through a disordered waveguide is determined by a self-consistent matrix transport equation for a matrix-valued radiance, derived from the microscopic wave equation. This 'radiant field theory' does not assume that thin slices scatter isotropically, so it stays accurate in the quasiballistic regime where the standard DMPK transport theory fails. The same equation accommodates absorption and partial control of the input channels, both common in experiments and previously lacking an ab initio treatment. Numerical solutions of the transport equation reproduce distributions computed from the microscopic wave equation in the regimes tested. If correct, this gives a single framework covering diffusive and quasiballistic transmission, with the diffusive DMPK result recovered as a direction-averaged limit.","feed_headline":"Matrix transport equation predicts wave transmission through disorder","feed_subtitle":"Derived from the wave equation, it covers diffusive and quasiballistic regimes, absorption, and partial aperture, matching simulations.","key_machinery":"The load-bearing object is the matrix-valued radiance $g(\\boldsymbol{\\Omega},\\mathbf{r})$, defined by a Wigner transform of the disorder-averaged Green's function projected on the wavenumber shell $\\|\\mathbf{p}\\|=k$. It obeys a first-order nonlinear transport equation in position and direction, with the normalized field $\\tilde Q(\\mathbf{r})=\\int d\\Omega\\, g(\\boldsymbol{\\Omega},\\mathbf{r})/S_d$ acting as a self-consistent scattering kernel, and with Heaviside factors $\\Theta(m_a-\\|\\boldsymbol{\\Omega}_\\perp\\|)$ that encode the numerical aperture. The boundary conditions (15) fix the incoming and outgoing angular sectors, and the constraints $g^2=1$, $\\operatorname{Tr}g=0$ replace the stronger isotropy normalization of $\\sigma$ models. This machinery converts the computation of $\\rho(T)$ from a many-channel statistical problem into the numerical solution of a few coupled transport equations, which is what makes absorption and partial aperture straightforward to include.","core_discovery":"The central claim is that the transmission eigenvalue distribution $\\rho(T)$ of a disordered waveguide is completely characterized by the matrix radiance $g(\\boldsymbol{\\Omega},\\mathbf{r})$, which solves the nonlinear transport equation (13) with boundary conditions (15). The derivation proceeds by averaging the wave equation over a Gaussian white-noise potential, using a replica trick and a saddle-point approximation, then passing to a Wigner (semiclassical) description; the result is a transport equation whose commutator structure encodes coherent interference. The paper validates the equation against large-scale numerical solutions of the microscopic wave equation in two-dimensional waveguides, including cases with $L/\\ell$ from 5 down to 0.05, with uniform absorption, and with reduced numerical aperture at the input. In the direction-averaged diffusive limit the equation reduces to the known DMPK result, and in the quasiballistic regime it agrees with numerics where DMPK deviates. The authors therefore claim a microscopic, parameter-free generalization of the standard theory that incorporates experimental realities such as absorption and incomplete channel control.","pith_inferences":["Since the boundary values of $g$ determine the generating function for $\\rho(T)$, an experiment that measures angle-resolved transmission could in principle reconstruct the matrix radiance itself, although the paper does not discuss such an inversion.","If the single-replica saddle-point approximation survives beyond the stated $L\\ll\\xi$ regime, the framework might describe the onset of Anderson localization, including the disappearance of the $T\\simeq1$ peak without a low-$T$ gap; the paper lists this as a promising direction but does not establish it.","In open geometries without waveguide side walls, the same $\\boldsymbol{\\Omega}\\cdot\\nabla_r$ transport structure could directly account for lateral diffusion, potentially replacing the fitted long-range correlations used in filtered random matrix theory with a direct calculation."],"forward_implications":["Transmission eigenvalue statistics in the quasiballistic regime ($L\\lesssim\\ell$) become computable from the wave equation, a regime where the standard DMPK theory is known to be inaccurate.","Absorption can be included by setting $\\varepsilon=k/\\ell_a$, covering the moderate-absorption range $\\ell_a\\le L\\lesssim\\xi_a$ that previously lacked an ab initio treatment.","Incomplete channel control enters through the aperture parameters $m_a,m_b$ in the transport equation, so rectangular transmission matrices are handled without the renormalized parameters of filtered random matrix theory.","The direction-averaged limit of the equation reproduces the DMPK result, showing the new theory as a strict generalization rather than a replacement.","Because the equation is derived from the microscopic Green's function, the same structure can be extended to other observables such as internal intensity and deposition, and to anisotropic or correlated disorder."],"supporting_citations":[{"why":"Supplies the DMPK theory and the bimodal transmission eigenvalue law that the new equation generalizes and the baseline it must match.","marker":"[5]"},{"why":"Documents the difficulty of including absorption in DMPK theory, the gap the new method fills.","marker":"[34]"},{"why":"Provides the filtered random matrix theory used as the comparison for incomplete channel control.","marker":"[35]"},{"why":"Companion paper containing the full derivation of Eq. (13), the boundary conditions, and the quasiballistic approximation.","marker":"[38]"},{"why":"Relates transmission matrix elements to the Green's function, the starting point of the field-theoretic computation.","marker":"[39]"},{"why":"The numerical solver for the RFT equations used to produce the theoretical curves.","marker":"[41]"},{"why":"The recursive Green's function code that generates the microscopic wave-equation distributions used for validation.","marker":"[42]"},{"why":"Provides the recursive Green's function method underlying the reference simulations.","marker":"[43]"}],"fun_headline_variants":["Beyond DMPK: transport equation for waves without isotropy","Matrix radiance transport predicts all-regime wave transmission","No-isotropy transport equation for coherent waves in disorder","Unified transport theory: absorption and partial aperture included"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole scheme depends on a saddle-point approximation in which replicas of the disordered system are treated as noninteracting; the authors state this is valid for samples shorter than the localization length but do not prove it or characterize how it fails.","fun_headline_variants_meta":{"raw":{"variants":["Beyond DMPK: transport equation for waves without isotropy","Matrix radiance transport predicts all-regime wave transmission","No-isotropy transport equation for coherent waves in disorder","Unified transport theory: absorption and partial aperture included"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001999,"raw_usage":{"total_tokens":7796,"prompt_tokens":939,"completion_tokens":6857,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":555,"completion_tokens_details":{"reasoning_tokens":6791}},"tokens_in":555,"tokens_out":6857,"duration_ms":42830,"temperature":1.0,"reasoning_tokens":6791,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T19:42:03.096741+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Solve Eq. (13) for a waveguide with optical thickness $L/\\ell$ fixed and gradually push the ratio $L/\\xi$ toward and beyond one by increasing disorder strength or shrinking the width, then compare with recursive Green's function simulations; if the RFT prediction departs from the microscopic distributions as localization is approached, the single-replica saddle-point approximation is the point of failure.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Documents the difficulty of including absorption in DMPK theory, the gap the new method fills."},{"cited_title":"Goetschy and A","cited_arxiv_id":null,"evidence_quote":"Provides the filtered random matrix theory used as the comparison for incomplete channel control."},{"cited_title":"Transmission eigenvalue distribution in disordered media from radiant field theory","cited_arxiv_id":"2411.10355","evidence_quote":"Companion paper containing the full derivation of Eq. (13), the boundary conditions, and the quasiballistic approximation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Relates transmission matrix elements to the Green's function, the starting point of the field-theoretic computation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The numerical solver for the RFT equations used to produce the theoretical curves."},{"cited_title":"MacKinnon and B","cited_arxiv_id":null,"evidence_quote":"Provides the recursive Green's function method underlying the reference simulations."}],"review_version":1}