{"id":"f0968f68-3dea-47c8-a371-f00a2375990c","arxiv_id":"2509.04725","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Two-photon correlations at a beam-splitter output can be spatially patterned and edited by tailoring the transverse polarization structure of the input photons.","lead":"The authors show that by structuring the polarization profile of two photons before they interfere at a beam splitter, they can write and edit the spatial correlation pattern between the two output ports. This lets high-dimensional information be encoded in photon correlations that stay invisible to ordinary intensity and single-port detection.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Predicted visibility maps rely on the untested assumption that both arms share the same transverse profile and z-independent polarization structure; no mode-matching check is reported.","rationale":"The reader's weakest assumption correctly identifies the hinge of the paper: the factorization of A(rC,rD) under the identical-transverse-profile assumption. All analytic visibility formulas, and the central claim of pointwise programmable correlations, depend on this. A separate internal inconsistency exists—the main text states V=(1/2)cos(φC−φD) for the circular-OAM case while the SI derives V=(1/2)cos2(φC−φD) (SI Eq. S20)—but this is a typo-level discrepancy that, once corrected, does not threaten the method. The factorization/Gouy concern is more load-bearing: if the two arms are not mode-matched, the predicted patterns wash out and the phrase 'write arbitrary spatially-dependent correlations' overstates the result. The experiment's qualitative agreement suggests the assumptions may hold in the tested configuration, but without single-arm profile and polarization characterization the robustness is unverified. Missing data/code and absence of error bars further justify a conditional rather than unconditional acceptance. I therefore see no reason to change the reader's CONDITIONAL verdict, but the concern is real and testable.","tokens_in":13214,"tokens_out":8294,"duration_ms":97569,"concrete_test":"With the same setup, block arm B and record the camera image from arm A alone; then block arm A and record arm B alone. Compute the normalized intensity overlap O = (∫I_A I_B d²r)² / (∫I_A² d²r ∫I_B² d²r) over the detection regions. Also make a spatially resolved Stokes measurement of each arm's polarization structure at the BS plane. Then recompute the predicted V(φC,φD) from SI Eqs. S11, S12, and S16 using these measured f_A, f_B, e_A, e_B rather than assuming identical profiles and z-independent modes. If O < ~0.99, or if the recomputed visibility map deviates from cos2φC cos2φD by more than the pixel-level shot-noise uncertainty, the factorization assumption is violated and the central claim must be qualified to mode-matched configurations.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central visibility formulas (SI Eqs. S18, S20, S29–S31) are derived from a coincidence probability that factorizes as A(rC,rD) times polarization overlaps. This factorization requires (i) the two input photons to have the identical transverse profile f(r,z,t) (SI, before Eq. S11) and (ii) the vector mode eσ(r) to be independent of z (main-text Eq. 4, SI Eq. S5). These assumptions let A cancel in V=(C_out−C_in)/C_out, leaving only polarization terms. In the experiment, each arm contains its own SMF, waveplate, q-plate, and free-space path to the BS; no imaging system is described that maps the q-plate planes to the BS or to the camera. Radial and π-VV modes are superpositions of OAM +1 and −1 polarization components, so free-space propagation rotates their polarization structure by the relative Gouy phase. If the two arms acquire different Gouy phases—or if their radial profiles, waists, or wavefront curvatures differ at the BS or at the camera—then e_A(rC) ≠ ideal radial/π and f_A(r) ≠ f_B(r). The factorization breaks, and the predicted maps, including the exact checkerboard V=cos2φC cos2φD and the exact zero visibility after integrating over one coordinate, are no longer valid. The paper's Discussion concedes this only qualitatively (“sensitivity to propagation-related Gouy phases … should be kept in mind”), and no single-arm spatial-profile or polarization-structure measurement is reported. Thus the pointwise correlation tailoring claimed as the central result is not yet established outside the specific, possibly fortuitously matched configuration.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper proposes and experimentally implements a method for tailoring the spatial correlations between two photons after a 50:50 beam splitter by structuring the transverse polarization profile (vector vortex modes) of the input photons. For temporally indistinguishable photons, the coincidence probability at output ports C and D is shown to depend pointwise on the local overlap of the structured modes; the paper derives parameter-free analytical visibility functions V=(C_out−C_in)/C_out for specific mode combinations, e.g., V=cos2φC cos2φD for radial and π VV modes. The experiment uses a TimePix3 event camera to record spatially resolved coincidences and demonstrates editing of the correlation maps by inserting polarizers. The authors claim this constitutes a scheme to write arbitrary spatially dependent correlations, with potential applications in high-dimensional quantum communication and imaging.","tokens_in":13514,"tokens_out":9455,"duration_ms":89493,"significance":"If fully validated, the scheme offers a simple, source-agnostic route to structuring biphoton correlations in the transverse plane, complementing SPDC-based approaches. The analytical treatment is self-contained, parameter-free, and yields falsifiable predictions. The experimental demonstration is visually compelling and the extension to multiphoton scenarios is plausible. However, the current manuscript does not provide quantitative agreement metrics, public data, or a direct check of the key mode-matching assumption, so the strength of the experimental support is substantially weaker than the text suggests. The idea is nevertheless novel and likely of interest to the quantum-optics community.","major_comments":[{"comment":"The visibility formulas (SI Eqs. S18, S20, S29–S31) are derived after factorizing A(rC,rD)=F(rC,zC)F(rD,zD), which requires the two input photons to share the same transverse profile f(r,z,t) and the vector mode eσ(r) to be independent of z. The experiment does not report any measurement of the single-arm spatial profiles or of the two-arm mode overlap at the BS and camera. Because radial and π-VV modes are superpositions of OAM ±1 components with polarization-dependent Gouy phases, differential Gouy phases or waist mismatches between the arms would break the factorization and alter the predicted maps, including the exact checkerboard and the zero visibility after single-coordinate integration. The Discussion only states qualitatively that Gouy-phase sensitivity 'should be kept in mind.' Please provide a direct mode-matching check (e.g., measured beam radii and wavefront curvatures at th","section":"SI, before Eq. (S11); main text Eq. (4) and Discussion"},{"comment":"No uncertainties, confidence intervals, or quantitative agreement metrics are reported for any experimental visibility map. The text repeatedly asserts 'good agreement' and acknowledges 'slight asymmetries' without numbers. Since the central claim is an experimental demonstration, the manuscript should include at least per-pixel error bars derived from Poissonian count statistics and a global fidelity or reduced chi-square between the measured and predicted V maps. The data-availability statement ('may be obtained from the authors upon reasonable request') falls short of the transparency expected for a claim of this strength.","section":"Figs. 2 and 3"},{"comment":"The main text states V(φC;φD) = 1/2 cos(φC − φD) for the radial-VV plus circular-OAM l=1 input, but the SI derivation gives V = 1/2 cos 2(φC − φD). The latter follows from Eq. (S11) for l=1; the main-text formula is inconsistent and changes the predicted periodicity. This is not merely a typo, because the subsequent sentence in the same paragraph uses the periodicity to discuss control by the OAM difference. Please correct the main-text formula and verify that the plotted theoretical map matches the SI expression.","section":"Main text, Results (third paragraph) vs SI Eq. (S20)"}],"minor_comments":[{"comment":"The phrase 'write arbitrary spatially-dependent correlations' overstates what is demonstrated: only two specific azimuthal mode families and a few projection settings are shown. Suggest softening to 'a class of' or 'tailored' spatial correlations.","section":"Introduction / Conclusion"},{"comment":"Please define the color scale, the meaning of φC and φD, and the radial integration procedure in the captions of Figs. 1–3, so that the maps are self-explanatory.","section":"Figure captions"},{"comment":"The expression for C_H,V(Out) appears to have a typographical artifact with a stray ' / 2' after 'sin^2 φD'. Please verify the factor.","section":"SI, Eq. (S27)"},{"comment":"Consider depositing processed visibility maps and coincidence-count data in a public repository instead of 'available upon request,' which would strengthen reproducibility.","section":"Data availability"}],"recommendation":"major_revision","confidential_remarks":"The theoretical framework is sound and the experiment is potentially interesting, but the missing quantitative validation and the unaddressed mode-matching assumption are load-bearing. The main-text formula typo for the OAM case should be fixed. I would not reject the paper; these issues can be addressed within the manuscript's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this paper deserves peer review. The idea is clean — spatially varying polarization to program pointwise HOM visibility between both output ports of a beamsplitter, plus polarization projections to edit the pattern. The theory is standard Glauber with explicit assumptions and no fitted parameters, and the chequerboard and stripe maps in Figs 2 and 3 look genuinely convincing. The observation that the structure is invisible to intensity measurements is a nice practical twist.\n\nThe main soft spot is internal inconsistency: the main text gives V = (1/2) cos(φC − φD) for the circular-OAM case, while the SI (Eq. S20) gives V = (1/2) cos 2(φC − φD). Those differ by a factor of two in angle; one is a typo or a wrong derivation. As written, a reader can't tell which one the experiment supports. That has to be fixed.\n\nThe experimental reporting is also thinner than I'd like: no error bars, no quantitative agreement metric, and the data are only 'available on request.' For a claim of arbitrary correlation shaping, that's not enough. The mode-matching issue the stress test raises is real, but I think it's secondary. The derivation assumes identical transverse profiles and z-independent polarization structure in both arms; the paper acknowledges Gouy-phase sensitivity only in passing. A referee should ask for a direct characterization of the modes at the detection plane. But the visibility maps themselves are a sensitive test: if the Gouy phases were badly mismatched, the chequerboard would wash out. So the fact that you see the pattern is indirect evidence the assumption holds. 'Indirect' is not a measurement, though.\n\nThe conclusion oversells: 'arbitrary' is not supported by azimuthal-only demonstrations, and 'secure communication' is speculation. If the authors fix the formula, add error bars and data release, and characterize mode matching, this is a solid experimental contribution. As is, it's a promising idea with an incomplete write-up.","headline":"A clean, parameter-free demonstration of pointwise HOM visibility shaping, but an internal formula discrepancy and thin experimental reporting keep it from being fully convincing as written.","tokens_in":14051,"tokens_out":4432,"would_cite":true,"duration_ms":45968,"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":"Two-photon correlation maps can be written point-by-point with structured polarization at a beam splitter.","keywords":["quantum interference","Hong-Ou-Mandel effect","vector vortex beams","spatial correlations","polarization structuring","coincidence imaging","high-dimensional quantum information","quantum eraser"],"falsifier":"Send two photons into the beam splitter with radial and π vector vortex modes, but deliberately give arm B a different radial profile (for example, a different Laguerre-Gaussian p index or a small defocus) while keeping the azimuthal polarization structure unchanged. If the factorized prediction V(φC, φD) = cos(2φC)cos(2φD) still fits the measured coincidence map, the shared-profile assumption is not needed at the claimed precision; if the visibility pattern washes out or develops a radial dependence, the assumption is falsified.","tokens_in":1675,"feed_emoji":"🌀","tokens_out":1919,"duration_ms":52182,"temperature":0.7,"pith_summary":"This paper introduces a way to tailor the spatial correlations of two photons after they interfere at a 50:50 beam splitter, by structuring each photon's polarization point-by-point across the transverse plane. The central claim is that local control of indistinguishability writes a spatially resolved quantum-interference pattern into the coincidence counts of the two output ports, even though ordinary intensity images show nothing. The authors derive analytical visibility functions for specific structured input modes, such as V(φC, φD) = cos(2φC)cos(2φD), and confirm them with an event-camera experiment using vector vortex modes. They also show that placing polarizers in the output ports edits these correlation patterns, adding a further layer of control. If correct, the scheme provides a simple route to encoding high-dimensional information in photon correlations for quantum communication and imaging.","feed_headline":"Structured light writes hidden two-photon correlation maps","feed_subtitle":"Patterning local polarization at a beam splitter tailors coincidence patterns that ordinary intensity images cannot see.","key_machinery":"The central object is the vector vortex mode, a transverse beam whose local polarization direction varies with azimuthal angle. These modes are prepared with q-plates and define a position-dependent polarization unit vector eσ(r) for each input photon. The beam-splitter transformation combines the two modes, and the coincidence probability between output positions is governed by the local overlap of the two polarization vectors: where they are locally indistinguishable, quantum interference suppresses or enhances coincidence counts; where they are locally distinguishable, it does not. The visibility function V(rC; rD) = (C_out − C_in)/C_out quantifies this pattern in a way that isolates the","core_discovery":"Two photons prepared in spatially varying polarization modes, sent into opposite ports of a 50:50 beam splitter, produce coincidence maps between the two outputs that change from point to point. For input modes that are the radial and π vector-vortex modes, the spatially resolved visibility takes the checkered form V(φC, φD) = cos(2φC)cos(2φD), alternating between bunching and anti-bunching with azimuthal period π. Because the two input modes are orthogonal, this structure is invisible to bucket detection: integrating over either output angle gives zero visibility. Replacing one input with a circularly polarized OAM mode changes the pattern to V(φC, φD) = ½cos(2(φC − φD)), and inserting pola","pith_inferences":["If the static q-plates were replaced by programmable spatial light modulators, the same mechanism could produce reconfigurable correlation patterns, turning the beam splitter into a programmable coincidence-pattern generator.","The factorization assumption underlying the analytical visibility formulas hinges on identical transverse profiles in both arms; experiments that deliberately introduce different radial profiles, wavefront curvatures, or Gouy-phase shifts could transform the visibility landscape in a predictable but uncharacterized way.","The pointwise distinguishability mechanism could be combined with orbital-angular-momentum sorting to build a high-dimensional encoding scheme where information is read only through coincidence measurements at matched spatial positions.","Because the same local-interference principle underlies both the 'hiding' and the 'editing' behavior, the scheme suggests a cryptographic primitive: a correlation key that is invisible under intensity detection and changes value when a projective polarization filter is inserted."],"forward_implications":["The written correlation patterns are invisible to ordinary intensity measurements, so high-dimensional information can be encoded in photon correlations and hidden from a classical observer.","For orthogonal input vector modes, spatial resolution in both output arms is required to see the correlations; bucket detection in either arm erases the structure entirely.","Polarization projections after the beam splitter edit the correlation maps, yielding uniform bunching, uniform anti-bunching, or single-coordinate patterns, which can serve as an additional decoding key.","The approach generalizes to more than two photons through Hong-Ou-Mandel interference, offering a route to multipartite structured correlations that avoids the scaling limits of spontaneous parametric down-conversion.","Local visibility maps reveal substantial spatial variation in two-photon interference, going beyond the global averages measured in standard Hong-Ou-Mandel experiments."],"supporting_citations":[{"why":"Provides the Hong-Ou-Mandel interference framework that the entire correlation-writing scheme builds on.","marker":"[23]"},{"why":"Demonstrates the spatially structured quantum eraser that this work extends from one output port to both output ports.","marker":"[29]"},{"why":"Supplies the q-plate technique used to prepare the radially varying vector vortex modes.","marker":"[19]"},{"why":"Defines the two-photon interference visibility function used throughout to quantify the correlations.","marker":"[36]"},{"why":"Gives the fourth-order correlation-function formalism that underlies the coincidence probability calculation.","marker":"[35]"},{"why":"Provides the quantum-eraser concept that motivates the post-beam-splitter polarization projections used to edit correlations.","marker":"[24]"},{"why":"Shows multiphoton Hong-Ou-Mandel interference with many input photons, supporting the paper's claim of scalability to multiparticle scenarios.","marker":"[39]"}],"fun_headline_variants":["Quantum interference sculpts hidden two-photon correlations","Beam splitter writes checkered coincidence maps invisible to intensity","Tailored polarization patterns hide bi-photon correlations","Spatial correlations shaped by quantum interference, not intensity","Invisible two-photon patterns from photon distinguishability"],"cache_read_input_tokens":15744,"weakest_assumption_plain":"The two input photons must share exactly the same transverse spatial profile, so that the coincidence probability factorizes into a product of a fluence envelope and a polarization-interference term; if the arms have different radial profiles, wavefront curvatures, or Gouy-phase-dependent structures, the predicted visibility maps break down.","fun_headline_variants_meta":{"raw":{"variants":["Quantum interference sculpts hidden two-photon correlations","Beam splitter writes checkered coincidence maps invisible to intensity","Tailored polarization patterns hide bi-photon correlations","Spatial correlations shaped by quantum interference, not intensity","Invisible two-photon patterns from photon distinguishability"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000861,"raw_usage":{"total_tokens":3519,"prompt_tokens":634,"completion_tokens":2885,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":378,"completion_tokens_details":{"reasoning_tokens":2808}},"tokens_in":378,"tokens_out":2885,"duration_ms":20610,"temperature":1.0,"reasoning_tokens":2808,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T05:55:48.597966+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Send two photons into the beam splitter with radial and π vector vortex modes, but deliberately give arm B a different radial profile (for example, a different Laguerre-Gaussian p index or a small defocus) while keeping the azimuthal polarization structure unchanged. If the factorized prediction V(φC, φD) = cos(2φC)cos(2φD) still fits the measured coincidence map, the shared-profile assumption is not needed at the claimed precision; if the visibility pattern washes out or develops a radial dependence, the assumption is falsified.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the Hong-Ou-Mandel interference framework that the entire correlation-writing scheme builds on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates the spatially structured quantum eraser that this work extends from one output port to both output ports."},{"cited_title":"Physical ReviewA94, 030304 (2016)","cited_arxiv_id":null,"evidence_quote":"Supplies the q-plate technique used to prepare the radially varying vector vortex modes."},{"cited_title":"& Zeilinger, A","cited_arxiv_id":null,"evidence_quote":"Defines the two-photon interference visibility function used throughout to quantify the correlations."},{"cited_title":"Cohen-Tannoudji, G","cited_arxiv_id":null,"evidence_quote":"Gives the fourth-order correlation-function formalism that underlies the coincidence probability calculation."},{"cited_title":"quantum eraser","cited_arxiv_id":null,"evidence_quote":"Provides the quantum-eraser concept that motivates the post-beam-splitter polarization projections used to edit correlations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows multiphoton Hong-Ou-Mandel interference with many input photons, supporting the paper's claim of scalability to multiparticle scenarios."}],"review_version":1}