{"id":"b3ce6643-34ba-4c53-8fb4-2286ff763ac3","arxiv_id":"2412.18244","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Thouless pumping of light is observed in continuous quasi-periodic photorefractive lattices, with the quantized beam displacement predicted by the Chern numbers of periodic approximants.","lead":"Light beams sent through a crystal with two overlapping, slowly sliding optical lattices travel sideways by a fixed amount each cycle. The shift is set by a topological number of the lattice, and the same number appears in a chain of periodic approximations to a truly quasi-periodic crystal.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The derivation of Eq. (2) rests on the unproved equal-population conjecture; the supporting text is internally inconsistent (equal populations imply many O(1/N) eigenvalues, not one dominant one), so the universal displacement formula needs a numerical density-matrix test.","rationale":"The reader's weakest assumption correctly identifies the equal-population conjecture as the load-bearing element of Eq. (2). My stress-test sharpens this: the paragraph supporting the conjecture is internally contradictory, because equal mini-band populations imply a density matrix proportional to the identity in that subspace, with roughly N equal eigenvalues, whereas the text claims one dominant eigenvalue. The formula Y_n = (L_n/N) C_m^1 requires the former reading; the rank-one reading would select a single mini-band Chern index and generically give a very different displacement. This is a correctness risk in the central theoretical derivation, not merely a question of matching experimental precision. The paper has real strengths: it demonstrates a clear experimental realization, the direction and approximate magnitude of the displacement match the predicted Chern numbers for two different irrational period ratios, and the convergence with the order of best rational approximants is shown. However, because Eq. (2) is the quantitative backbone of the universality claim and its derivation is not self-consistent as written, the result should remain conditional pending a direct numerical check of the density-matrix structure. The post-hoc 20% intensity threshold in the experimental comparison further reinforces the need for raw data or an independent analysis. I therefore keep the reader's CONDITIONAL verdict unchanged.","tokens_in":13474,"tokens_out":6831,"duration_ms":65752,"concrete_test":"For the same parameters as Figs. 3 and 4, numerically solve Eq. (1) in the 4th and 5th BRAs for both irrational periods, using several input beam widths, positions, and V0 values. At z=Z, expand the output field in the instantaneous Bloch basis of the n-th approximant, form the N x N density matrix rho, compute its eigenvalues, and evaluate Tr(rho C). If the eigenvalues are not all close to 1/N (i.e., if one is O(1) and the rest negligible) or if Tr(rho C) deviates from C_m^1/N by more than about 5%, Eq. (2) fails and the universal-displacement claim requires revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim, Eq. (2), is not derived from first principles: it is introduced as a conjecture ('one may conjecture that in the course of sufficiently long evolution the populations of the mini-bands become nearly equal', text near Eq. (2)). The justification is internally inconsistent. If the populations of the N_m^n mini-bands are nearly equal, the normalized density matrix is approximately (1/N_m^n) I, so its eigenvalues are all approximately 1/N_m^n, not 'one eigenvalue of order one and all other eigenvalues negligible' as the same paragraph states. The formula Y_n(Z) = (L_n/N_m^n) C_m^1 follows from the uniform-density-matrix reading (because Tr(rho C) = C_m^1/N_m^n); a rank-one density matrix would instead give the Chern index of a single mini-band, which generically takes values such as (-1)^n q_{n-1} or (-1)^n(q_{n-1}-q_n) and would not equal the average. Thus, as written, the derivation of Eq. (2) is not self-consistent, and the claimed universality with respect to input beam shape and lattice depth is not established. The experimental agreement in Figs. 3D and 4D is partly obtained only after a 20% intensity threshold ('experiment 2'), and raw data and code are not released, so the experiment does not by itself resolve the conjecture.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports an experimental study of Thouless pumping of light in a continuous photorefractive quasi-periodic potential, realized by superimposing two periodic sublattices with incommensurate periods and emulated by periodic approximants (best rational approximations, BRAs). The central theoretical claim, expressed in Eq. (2), is that after one pumping cycle the center-of-mass displacement of a paraxial beam in the quasi-periodic limit is Y_phi(Z) = L_phi C_m^1, where C_m^1 is the Chern number of the highest band of the m-th approximant at which pumping is quasi-adiabatic and L_phi is the limiting ratio of the approximant period to the number of mini-bands. The authors report positive and negative displacements for two irrational period ratios, consistent with the sign of the predicted Chern numbers, and observe saturation of the displacement with increasing approximant order. They further argue that the result is universal, independent of the input beam shape and lattice depth.","tokens_in":13729,"tokens_out":4751,"duration_ms":41626,"significance":"If the central formula (2) is established, the paper would provide a significant experimental demonstration of topological pumping in a genuinely continuous quasi-periodic system, connecting transport to the Chern numbers of periodic approximants and introducing a conceptually useful 'quasi-adiabatic' regime. The sign reversal of the displacement between the two lattices is a strong qualitative control, and the observed saturation with BRA order supports the approximant-based picture. However, the quantitative prediction relies on an unproved and internally inconsistent conjecture about the density matrix of mini-band populations, and the agreement with experiment is partly obtained through an ad hoc intensity threshold. The experimental observation is valuable, but the theoretical foundation of the universal quantitative claim needs to be repaired.","major_comments":[{"comment":"The derivation of Eq. (2) rests on a conjecture that is internally inconsistent. The paragraph states that 'in the course of sufficiently long evolution the populations of the mini-bands become nearly equal', but then asserts that the density matrix at the output 'has one eigenvalue of order one and all other eigenvalues negligible'. If the populations are nearly equal across N_m^n mini-bands, the normalized density matrix is approximately the identity divided by N_m^n, so all eigenvalues are approximately 1/N_m^n. Under that uniform reading, Tr(rho C) equals C_m^1/N_m^n and Eq. (2) follows; under the rank-one reading, the displacement would instead equal the Chern index of a single mini-band, which generically takes values such as (-1)^n q_{n-1} or (-1)^n(q_{n-1}-q_n), not the average. As written, the derivation of Eq. (2) is not self-consistent. Please provide a numerical computation of the eigenvalue distribution of the density matrix at z=Z for the approximants in Figs. 3 and 4, and verify directly that Tr(rho C) equals the right-hand side of Eq. (2), or provide a corrected analytical argument. Without this, the claimed universality with respect to input beam shape and lattice depth is not established.","section":"Text near Eq. (2), 'To characterize the observed pumping quantitatively'"},{"comment":"The quasi-adiabatic order m is a free input in the quantitative comparison with experiment. The definition is qualitative: transitions between the highest band and lower bands of the m-th approximant are suppressed, while transitions between mini-bands of the (m+1)-th approximant occur. For the first experiment m=1 is selected, and for the second m=2 is selected based on the band-structure analysis in Fig. S5. Because m determines which Chern number appears in Eq. (2), the paper should provide a systematic, quantitative criterion for m (for example, Landau-Zener estimates of the transition probabilities using the computed gaps and the experimental velocity v) and a sensitivity analysis showing how the predicted displacement changes for m=1, 2, and 3 in both experimental configurations. As it stands, the choice of m could be considered a parameter selected a posteriori to match the observed sign and magnitude.","section":"Paragraph beginning 'We say that a quasi-adiabatic condition is satisfied...'"},{"comment":"The quantitative agreement between the measured displacement and Eq. (2) is obtained only after discarding the field below 20% of the peak intensity. The raw center-of-mass ('experiment 1') deviates from the prediction by an amount that is significant relative to the effect size, especially in Fig. 4D. The intensity cutoff is introduced as a post-hoc adjustment ('by disregarding a certain amount of light side lobes') and no independent justification is given. Because the central claim is the quantitative value of the displacement, please show the computed displacement as a function of the intensity threshold for both experiments, and demonstrate a plateau or an objective criterion (e.g., the spatial extent of the localized guided modes) that selects the 20% threshold. Otherwise the agreement in Figs. 3D and 4D is not a reliable test of Eq. (2).","section":"Figs. 3D and 4D, 'experiment 2'"}],"minor_comments":[{"comment":"The phrase 'in physical media where it occurs' is awkward; consider rephrasing for clarity, and the misspelling 'Schr¨ odnger' in the Introduction should be corrected to 'Schrödinger'.","section":"Abstract and Introduction"},{"comment":"The ordinate labels differ between Fig. 3B ('Y (mm)') and Fig. 3D ('Y_n (mm)'), and similarly for Fig. 4; please use consistent notation and clarify the units (the '×10^-2' scaling is confusing when read with the axis tick labels).","section":"Fig. 3 and Fig. 4 captions and axes"},{"comment":"The sign convention in Eq. (5c) for the wavevector of pinhole 3, specifically the factor (1 - 2 p_n/q_n), is not explained relative to the description of the pinhole positions; please add a sentence clarifying the geometry.","section":"Methods, Eqs. (5) and (6)"},{"comment":"The data availability statement says that data are 'available from the corresponding author upon reasonable request' but no repository or persistent identifier is provided. For a quantitative experimental paper, releasing the raw COM trajectories and analysis scripts would strengthen reproducibility.","section":"Data availability"},{"comment":"In Fig. 2B, the caption does not specify which color corresponds to which BRA order; please add explicit color/line-style definitions.","section":"Fig. 2 caption"}],"recommendation":"major_revision","confidential_remarks":"This is a potentially important experimental result, and the sign reversal between the two quasi-periodic lattices is a strong control. However, the theory underlying the quantitative formula (2) is not self-consistent as written: the equal-population conjecture conflicts with the stated one-dominant-eigenvalue property. This is fixable with a numerical density-matrix test, which the authors likely already have the tools to perform. The ad hoc 20% intensity cutoff also needs justification. I would not reject the paper, but it should not be accepted in its current form. I also note that the authors choose m=1 and m=2 in the two experiments, and a more rigorous definition of quasi-adiabatic order would substantially increase confidence."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core result is genuinely new: the first observation of quasi-adiabatic Thouless pumping of light in a continuous photorefractive quasiperiodic lattice, emulated by periodic approximants. That is worth taking seriously. The experiments are also well constructed—two different irrational period ratios, a sign-reversing control lattice, and a clear saturation of the displacement with increasing approximant order. The direction and rough magnitude of the center-of-mass shift track the Chern numbers of the relevant approximant bands, and the sign reversal for the second lattice is a strong piece of evidence that the effect is topological.\n\nThe soft spot is the theoretical derivation of Eq. (2), the universal displacement formula. The text says that after long evolution the mini-band populations become nearly equal, and then immediately says the density matrix has one eigenvalue of order one with all others negligible. Those two statements contradict each other: nearly equal populations would give a density matrix close to the identity, with all eigenvalues roughly 1/N. So the derivation of Eq. (2) as written is not self-consistent. The formula may still be right, but the justification is not, and the claimed universality with respect to input beam shape and lattice depth is not established by the paper's argument. This is not a minor typo; it is the step that connects the measured displacement to the Chern number.\n\nThere are also smaller issues. The reported agreement for the second lattice is improved by a 20% intensity cutoff in the COM analysis, which is a post-hoc processing choice, and raw data and code are not released. The choice of which BRA order m is “quasi-adiabatic” is made from band-structure analysis, so there is some freedom in selecting which Chern number to compare, though the direction and sign reversal are not fitted.\n\nNone of this kills the paper. The experimental observation is solid and the qualitative connection to topology is convincing. But the quantitative claim in Eq. (2) needs a rigorous numerical test of the density-matrix evolution, not another conjecture. The authors should also release the raw COM data and settle the eigenvalue inconsistency.\n\nI would send this to peer review. It deserves a serious referee, but with clear requests: fix the density-matrix argument, test the conjecture numerically, and make the data available. The audience is topological photonics and quasiperiodic systems researchers; for them this is an important step.","headline":"A real experimental first—Thouless pumping in a continuous quasiperiodic lattice—but the theory behind the universal displacement formula is internally inconsistent and needs a numerical fix before the quantitative claim holds.","tokens_in":14296,"tokens_out":1529,"would_cite":true,"duration_ms":16278,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A light beam in a continuous quasi-periodic photorefractive crystal shifts by a quantized distance per pumping cycle, the value set only by the sublattice period ratio and the Chern number of the approximant band that carries the pumping.","keywords":["Thouless pumping","quasi-periodic photonic crystal","periodic approximants","best rational approximations","Chern number","photorefractive crystal","quasi-adiabatic pumping","mobility edge"],"falsifier":"Launch the input beam into a single high mini-band of a fixed approximant and measure the one-cycle center-of-mass displacement: if the equal-population conjecture holds, the displacement must approach $L_\\varphi C_m^1$ regardless of which mini-band is excited (after enough propagation), whereas a systematic dependence on the chosen mini-band would falsify Eq. (2). A complementary numerical check: compute the eigenvalues of the output density matrix $\\rho_n(Z)$ over a pumping cycle — the conjecture requires one eigenvalue of order one with all others negligible.","tokens_in":13228,"feed_emoji":"💡","tokens_out":14056,"duration_ms":113013,"temperature":0.7,"pith_summary":"The paper reports the experimental observation of Thouless pumping — quantized topological transport — of a light beam in a genuinely continuous, incommensurate photorefractive crystal, a regime in which the standard notion of adiabaticity fails because the spectrum is dense at every pumping velocity. The authors emulate the quasi-periodic potential with periodic approximants built from the best rational approximations of the irrational ratio between the two sublattice periods, and claim that the one-cycle displacement of the beam's center of mass equals $L_\\varphi C_m^1$: the limiting ratio of the approximant period to its number of mini-bands, multiplied by the Chern number of the highest band of the approximant at which pumping is quasi-adiabatic. If the claim is right, the quantized shift is universal — fixed only by the period ratio, the sliding angle, and the Chern number, and independent of the beam shape, of which mini-band is excited below the mobility edge, and of the lattice depth. The result would carry topological pumping from periodic and discrete quasi-periodic systems into continuous quasi-crystals, and would make the continued-fraction structure of an irrational number directly visible in a tabletop optics experiment.","feed_headline":"Light shifts by quantized steps in a quasiperiodic crystal","feed_subtitle":"One-cycle beam displacement is fixed by the sublattice period ratio and a Chern number, verified in a 2-cm continuous lattice.","key_machinery":"The carrying object is the family of periodic approximants $H_n$ of the quasi-periodic Hamiltonian, obtained by replacing the irrational period ratio $\\varphi$ with its best rational approximations $p_n/q_n$ — the convergents of its continued fraction. Each approximant is periodic in the transverse coordinate with period $L_n=\\pi q_n$ and in the propagation (time-like) coordinate with period $Z$, so its bands carry conventional Chern numbers on the torus $[0,L_n)\\times[0,Z)$. The one-cycle displacement is written as $Y_n(Z)=L_n\\,\\mathrm{Tr}\\{\\rho_n C_n\\}$, the trace of the output density matrix $\\rho_n$ against the Chern matrix $C_n$; under the equal-population conjecture this reduces to $Y_\\varphi(Z)=L_\\varphi C_m^1$. Quasi-adiabaticity at order $m$ means that for the chosen velocity, transitions between the highest band of $H_m$ and lower bands are suppressed while transitions between the mini-bands that emerge at order $m+1$ are not. Two mechanisms carry the argument to the experiment: the memory effect, under which bands of lower approximants persist unchanged in higher ones with Chern numbers that encode the continued-fraction history, and the mobility edge, which keeps all excited modes localized inside a single period so the beam never distinguishes the approximant from the true quasi-periodic potential.","core_discovery":"The central claim is that quasi-adiabatic Thouless pumping occurs in a genuinely continuous incommensurate photorefractive lattice — emulated by periodic approximants built from the best rational approximations of the irrational period ratio $\\varphi$ — and that the one-cycle center-of-mass displacement of the beam converges, in the quasi-periodic limit, to $Y_\\varphi(Z)=L_\\varphi C_m^1$, where $L_\\varphi=\\lim_{n\\to\\infty}L_n/N_n^m$ is the limiting ratio of the $n$-th approximant's transverse period $L_n=\\pi q_n$ to the number $N_n^m$ of mini-bands into which the highest band of the $m$-th approximant splits, and $C_m^1$ is the Chern number of that highest band computed in the $(y,z)$ torus. For $\\varphi=1/\\sqrt{5}$ the prediction is $Y_\\varphi(Z)=\\pi\\sqrt{5}$ with $C_1^1=1$; for $\\varphi=(\\sqrt{5}+1)/4$ it is $Y_\\varphi(Z)=\\pi/(1-\\varphi)$ with $C_1^2=-1$, so the beam is pumped opposite to the sliding sublattice. The measured displacements agree with these values after accounting for radiation loss and background noise, and they saturate already at the 3rd–4th approximant, showing that the infinite quasi-periodic limit is effectively reached in a 2-cm sample.","pith_inferences":["If the equal-population conjecture holds, the same formula should transfer to any continuous one-dimensional incommensurate potential with a mobility edge; a scan over irrational period ratios $\\varphi$ would test whether the displacement tracks $\\pi\\lim_n q_n/N_n^m$ with the splitting rule $N_n^m$ dictated by the continued fraction.","The scheme implies a velocity staircase: as the pumping velocity is reduced stepwise, successive approximants become quasi-adiabatic and the output position should jump through a sequence of quantized values indexed by the convergents $p_n/q_n$ — a direct physical readout of the continued-fraction expansion of $\\varphi$.","A diagnostic of the conjecture is the eigenvalue spectrum of the output density matrix: one dominant eigenvalue confirms the near-equal mini-band populations, whereas several comparable eigenvalues would predict precisely how the displacement depends on the input beam's overlap with individual mini-bands.","The authors' closing suggestion of two- and three-dimensional moiré lattices implies an untested generalization: the pumping direction and magnitude would be set by the full vector of period ratios and sliding angles, enabling topologically controlled routing of wavepackets in higher-dimensional aperiodic settings."],"forward_implications":["The one-cycle displacement of a beam in a continuous quasi-periodic potential equals $L_\\varphi C_m^1$, with $L_\\varphi$ fixed by the sublattice period ratio: $+\\pi\\sqrt{5}$ for $\\varphi=1/\\sqrt{5}$ (Chern number $+1$) and $\\pi/(1-\\varphi)$ with Chern number $-1$ for $\\varphi=(\\sqrt{5}+1)/4$, the latter beam moving opposite to the sliding sublattice.","The shift is independent of the input beam shape, of which mini-band below the mobility edge is excited, and of the lattice depth $V_0$; only the period ratio, the sliding angle, and the Chern number enter.","The quasi-periodic limit is reached already at low approximant orders — saturation is visible by the 3rd–4th best rational approximation — so the prediction is testable in a 2-cm photorefractive sample.","Genuine quasi-periodic media are never adiabatic in the standard sense at any velocity; the observed quantization requires the paper's quasi-adiabatic definition tied to a specific best-rational-approximation order $m$, and lowering the velocity shifts $m$ and with it the Chern number and the output position."],"supporting_citations":[{"why":"defines quantized adiabatic transport (Thouless pumping) in periodic potentials, the phenomenon this work extends to continuous quasi-periodic media","marker":"[1]"},{"why":"demonstrates earlier quasi-periodic pumping in discrete waveguide arrays, the baseline this continuous-medium experiment goes beyond","marker":"[3]"},{"why":"introduces the periodic-approximant method for quasi-periodic potentials that the experiments emulate","marker":"[21]"},{"why":"documents the memory effect in approximants of quasi-periodic media, the mechanism by which higher-order approximations leave low-order dynamics unchanged","marker":"[24]"},{"why":"supplies the continued-fraction theory whose convergents define the best rational approximations used to build each approximant","marker":"[27]"},{"why":"provides the photorefractive optical-induction model used in the governing equation for the light propagation","marker":"[28]"},{"why":"establishes optically induced photonic lattices as the experimental platform for the measurements","marker":"[29]"},{"why":"defines the mobility edge that separates the localized modes keeping the beam confined during pumping","marker":"[38]"}],"fun_headline_variants":["Chern number sets light's step in quasi-crystal","Golden ratio dictates quantized light pumping","Continuous quasi-crystal shows Thouless pumping","Beam displacement quantized by band Chern number","Quasiperiodic lattice pumps light with integer steps"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The universal value of the displacement rests on the conjecture that over a sufficiently long pumping cycle the mini-band populations become nearly equal, so the output state has one dominant eigencomponent; without that equalization the shift depends on which mini-bands the input beam excited, and the choice of approximant order $m$ at which quasi-adiabaticity is declared selects which Chern number the data are judged against.","fun_headline_variants_meta":{"raw":{"variants":["Chern number sets light's step in quasi-crystal","Golden ratio dictates quantized light pumping","Continuous quasi-crystal shows Thouless pumping","Beam displacement quantized by band Chern number","Quasiperiodic lattice pumps light with integer steps"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000363,"raw_usage":{"total_tokens":2003,"prompt_tokens":1040,"completion_tokens":963,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":656,"completion_tokens_details":{"reasoning_tokens":892}},"tokens_in":656,"tokens_out":963,"duration_ms":8839,"temperature":1.0,"reasoning_tokens":892,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T04:53:23.220993+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Launch the input beam into a single high mini-band of a fixed approximant and measure the one-cycle center-of-mass displacement: if the equal-population conjecture holds, the displacement must approach $L_\\varphi C_m^1$ regardless of which mini-band is excited (after enough propagation), whereas a systematic dependence on the chosen mini-band would falsify Eq. (2). A complementary numerical check: compute the eigenvalues of the output density matrix $\\rho_n(Z)$ over a pumping cycle — the conjecture requires one eigenvalue of order one with all others negligible.","supporting_citations":[{"cited_title":"Thouless, Physical Review B 27, 6083 (1983)","cited_arxiv_id":null,"evidence_quote":"defines quantized adiabatic transport (Thouless pumping) in periodic potentials, the phenomenon this work extends to continuous quasi-periodic media"},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"demonstrates earlier quasi-periodic pumping in discrete waveguide arrays, the baseline this continuous-medium experiment goes beyond"},{"cited_title":"Tanese, E","cited_arxiv_id":null,"evidence_quote":"introduces the periodic-approximant method for quasi-periodic potentials that the experiments emulate"},{"cited_title":"Modugno, New Journal of Physics 11, 033023 (2009)","cited_arxiv_id":null,"evidence_quote":"documents the memory effect in approximants of quasi-periodic media, the mechanism by which higher-order approximations leave low-order dynamics unchanged"},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the continued-fraction theory whose convergents define the best rational approximations used to build each approximant"},{"cited_title":"Bohr, Acta Mathematica 45, 29 (1925)","cited_arxiv_id":null,"evidence_quote":"provides the photorefractive optical-induction model used in the governing equation for the light propagation"},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"establishes optically induced photonic lattices as the experimental platform for the measurements"},{"cited_title":"Biddle and S","cited_arxiv_id":null,"evidence_quote":"defines the mobility edge that separates the localized modes keeping the beam confined during pumping"}],"review_version":1}