Pith. sign in

REVIEW 4 major objections 5 minor 56 references

Integrated Photonic Programmable Random Matrix Generator with Minimal Active Components

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper claims that two programmable phase layers, separated by a fixed dense waveguide mixing lattice, turn sparse optical inputs into outputs that pass white-noise tests.

desk verdict Two-layer whitening claim is plausible and worth refereeing, but the intensity-only experiment cannot actually test the second phase layer and the general sufficiency claim outruns the evidence. read the letter →

arxiv 2501.08953 v1 pith:VF2NTBL6 submitted 2025-01-15 physics.optics physics.app-phquant-ph

Kevin Zelaya , Mostafa Honari-Latifpour , Mohammad-Ali Miri This is my paper · ORCID
classification physics.opticsphysics.app-phquant-ph
keywords randommatrixgenerationphotonicintegratedcircuitsprogrammablephaseshifterswhitenoiseopticalencryptionwaveguidelatticesinterlacedarchitecturessiliconphotonics
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper claims that a programmable photonic circuit can generate random unitary transformations using only two programmable phase layers, regardless of how sparse or structured the input is. It argues that two layers of random phase shifts, separated by a fixed waveguide lattice that mixes all channels, are enough to make the output statistically indistinguishable from white noise. The authors support the claim with numerical simulations at $N=100$ and a silicon-photonics experiment at $N=5$, in which sparse binary inputs were demultiplexed, randomized, and read out as power with the Rayleigh statistics expected for complex white noise. If the claim holds, compact on-chip random matrices become practical for optical encryption, random projections, and related photonic information tasks.

What carries the argument

The central object is the interlaced unitary $U = P^{(M)} F(\alpha_M) \cdots P^{(1)} F(\alpha_1)$, where each $P^{(j)}$ is a diagonal layer of random phase shifts $e^{i\phi_n^{(j)}}$ and $F(\alpha)=e^{i\alpha H}$ is a fixed passive mixing layer built from a waveguide lattice. The mixing layer must satisfy a 'density criterion': its transfer matrix must be dense enough to shuffle every input channel into all output channels. The argument then runs on two white-noise criteria, truncated autocorrelation and Shannon entropy applied to the output modulus, with Rayleigh statistics for power measurements, to show that $M=2$ already reaches the white-noise region and $M>2$ adds no significant improvement.

What would settle it

Search numerically over random dense unitary mixing matrices $F$ at $N=100$: if any $F$ that satisfies the paper's density criterion leaves the $M=2$ output entropy below the white-noise band for a sparse input, the general two-layer claim would be disproved. A simpler version is to test the two-layer device with a deliberately constructed mixing matrix that is dense but highly structured, such as a block-diagonal unitary with dense blocks, and check whether a sparse input remains localized.

Watch

Extended reading notes

Core claim

Supported by numerical simulation at $N=100$ and by a fabricated 5-port silicon photonic circuit, the paper claims that random phase masks in just two layers, $M=2$, interlaced with fixed dense waveguide-lattice mixers, are sufficient to convert arbitrary sparse input vectors into outputs that behave as white noise. The transfer matrix is $U = P^{(2)} F P^{(1)} F$, with $P^{(j)}$ random diagonal phases and $F=e^{i\alpha H}$ a fixed unitary mixing layer. With one layer, outputs retain structure and entropy below the white-noise band; adding the second layer washes out that structure, and a third or fourth layer yields no significant improvement. Because $U$ is unitary, the same device can decrypt by injecting the encrypted signal into $U^{\dagger}$ with the conjugate phases, and the paper experimentally verifies the randomization using power measurements, whose Rayleigh statistics flag complex white noise.

Load-bearing premise

The claim depends on the assumption that the fixed mixing layer is 'dense enough' to mix all channels; this density criterion is qualitative and has only been tested on two specific lattices, so a different mixing layer that looks dense could in principle fail to make two phase layers sufficient.

Editorial extensions

If this is right

  • A programmable random unitary can be built with $2N$ phase shifters plus one fixed lattice, instead of the $O(N^2)$ phase shifters of general Mach-Zehnder meshes.
  • For uniformly distributed phases the output randomness is stronger than for normally distributed phases, so key design can choose between noise quality and security.
  • The device can encrypt a 100-dimensional sparse signal by demultiplexing it into 5-dimensional sequences; the same physical chip then decrypts by running the inverse unitary.
  • Adding random deformation to the passive layer changes the effective transfer matrix, so the same architecture can emulate random or disordered waveguide systems with tunable disorder.
  • Requiring both entropy and truncated-autocorrelation criteria rules out the false positive of a single pulse, giving a practical test for whether an optical signal is white noise.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural next step is to test whether the $M=2$ device produces not only white-noise outputs but unitaries close to Haar-random; the paper does not address this, but the compact architecture would be an attractive candidate for boson sampling if it does.
  • The density criterion could be made quantitative, for example by a lower bound on the mixing matrix's ability to spread a single channel over all outputs, which would let designers certify new lattice geometries without full simulations.
  • Because the experimental readout is power-only, the paper's randomness evidence does not cover the phase of the complex output; a phase-resolved version would strengthen the claim and is directly enabled by the maintained unitarity.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes a compact programmable photonic circuit for random matrix generation, consisting of two random phase layers interlaced with fixed dense mixing layers (waveguide lattices). The central claim, stated in the abstract and Section 2D, is that M=2 random phase layers suffice to whiten arbitrary sparse input signals, as measured by entropy and autocorrelation criteria against white-noise ensembles. The authors support this with numerical simulations at N=100 for sparse pulse inputs, an experimental demonstration on a 5-port silicon photonics chip using power measurements, and an application to all-optical image encryption. The experimental section explicitly notes that the second phase layer is invisible in power readouts.

Significance. If the two-layer sufficiency claim holds rigorously, the architecture would substantially reduce the number of active controls needed for programmable random unitary operations in photonic information processing, with potential impact on optical encryption, random projections, and boson-sampling-related tasks. The paper's strengths include a transparent system model in Eq. (1), numerical tests on a nontrivial sparse-input ensemble at N=100, consideration of two statistical criteria to reduce false positives, and a concrete fabricated 5-port chip with packaged control. The encryption demonstration with perturbed mixing layers is an interesting additional result. However, the central claim is currently supported only by a permissive statistical acceptance rule and by numerical simulation for two specific mixing lattices, and the experiment does not actually test the two-layer complex-field transform because power detection removes the second phase layer. These gaps are load-bearing for the abstract's generality claim.

major comments (4)
  1. [Section 2D, Fig. 4e-f] The numerical evidence for two-layer sufficiency rests on an acceptance criterion that requires the entropy and truncated-autocorrelation statistics to fall within the mean ± 1σ band of white-noise ensembles. This is not a hypothesis test with controlled error rates. The text states that for M=2 'only a few' of the 50 samples lie outside the shaded region, but it does not report the actual count or pass rate for either criterion. A reader cannot determine whether the observed excursions are consistent with sampling fluctuation or indicate systematic failures for certain sparse inputs. Please report the exact number of accepted samples and, ideally, a formal goodness-of-fit or significance test against the null hypothesis of white noise.
  2. [Section 2D and Eq. (1)] The experimental power measurements cannot validate the two-layer sufficiency claim. Since the output is detected as |Ux|^2 and P(2) is a diagonal phase matrix, we have |Ux|^2 = |F P(1) F x|^2, so the second phase layer P(2) is completely erased. The measured Rayleigh-like statistics therefore demonstrate only that one random phase layer between two fixed mixers produces a whitened intensity profile, not that two layers are sufficient. The abstract's statement 'We experimentally demonstrate these results' is thus an overstatement. Either phase-resolved measurements (e.g., interferometry or an optical vector analyzer) are needed to test the two-layer claim, or the experimental claims should be explicitly limited to the effective one-layer intensity transform.
  3. [Section 2A and Discussion] The general claim that two layers suffice 'for any random lattice that fulfills the density criterion' is unsupported. The density criterion from ref. [26] is qualitative, and the manuscript tests only the Jx and homogeneous lattices numerically at N=100 and one lattice experimentally at N=5. Section 2E's perturbation study varies the passive layer randomly, but it reports only decryption fidelity of an image, not whether the whitening criteria are met for sparse inputs under Eq. (3). To support the universal phrasing, the authors should either provide a proof or a rigorous numerical test over an ensemble of dense random unitary passive layers (e.g., Haar-random F) checking the same whitening criteria for sparse inputs.
  4. [Section 2D, Fig. 4e-f; 'even for highly sparse input'] The set of 50 sparse inputs used in the numerical study consists of random placements of unit pulses with at least one pulse per demultiplexed 5-dimensional block. This is a limited family of sparse vectors; it does not establish whitening for 'arbitrary sparse inputs' as stated in the abstract and Section 2D. For example, inputs with clusters, with all zeros in some blocks, or with unequal pulse amplitudes are not tested. Please either narrow the claim to the tested family or expand the numerical experiments to a broader class of sparse signals.
minor comments (5)
  1. [Fig. 4 caption] The caption refers to 'Fig. 4g' for the real part of the transfer matrices and then again to 'Fig. 4g' for the real and imaginary parts of the first 25 encrypted samples; the figure panel labels appear inconsistent. Please renumber the panels so that each referenced panel is unique.
  2. [Section 2C, Eq. (2)] The definition of the truncated autocorrelation Ẋx[x] as (X_1,...,X_{N-1}) is given, but the later definition of the finite difference Δ_ℓ Ẋx[|x|] := (X_2 - X_1, ..., X_{N-2} - X_{N-1}) uses X_0 implicitly for the first term; please clarify the indexing convention so that the lag-ℓ correspondence is unambiguous.
  3. [Methods, entropy estimation] The number of histogram bins M = ⌊√N⌋ is stated, but no discussion is given of how the bin edges are chosen for signals of different dynamic ranges (e.g., sparse inputs versus Rayleigh-distributed powers). Since the entropy value depends on binning, please state the binning rule and briefly justify its stability.
  4. [Throughout] There are typographical and grammatical issues that should be corrected, for example 'randomly placed unit pulses' in Section 2D, 'Decelerations' in the declaration line, and the use of 'Eq. Eq. (1)' in the Fig. 4 caption.
  5. [Section 2B] The text says 'the PIC output is gathered through a power meter, any phase information is washed out during power measurements' and then later notes the second layer does not modify the power readout. These statements are correct but should be placed together near the experimental design discussion, so that the reader immediately understands the experimental scope.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the two-layer sufficiency claim is supported by direct numerical/experimental evaluation against externally defined white-noise benchmarks; the density-criterion self-citation is a generalization premise, not a fitted input, and the power-measurement blind spot is a stated experimental limitation rather than a circular derivation.

full rationale

The paper's central claim—that M=2 random phase layers suffice to whiten even sparse inputs—is supported by direct simulation (Section 2D, Fig. 4) and by experiment (Section 2B), not derived from a fitted parameter or from the density criterion. The white-noise benchmarks (normal and Rayleigh ensembles) in Section 2C are computed from 1000 synthetic signals per N before the device outputs are examined, so the acceptance regions are external to the device and not fitted to its outputs. The numerical sweep M=1,2,3,4 directly compares the architecture at different layer counts, and the claim that M=2 is sufficient is an empirical observation from those simulations rather than an equation forced by construction. The self-citation to ref. [26] (density criterion/Goldilocks principle) is used to justify that the passive layer F can be any sufficiently dense matrix, but the demonstrated results use specific Jx and homogeneous lattices; the density criterion is a generalization premise, not the source of the observed output statistics, so it is not a load-bearing circular step. One admitted limitation is that power measurements are insensitive to the final phase layer P^(2) (Section 2D: 'for power measurements, the second phase layer in the architecture in Figure 1 does not modify the readout at the power detectors'), so the experiment validates at most one phase layer between two mixers; however, this is a measurement limitation explicitly acknowledged by the authors, not a case of the prediction being defined by the fit. The statistical acceptance rule (mean ± 1 sigma bands) is permissive and could be criticized as a weak test, but that is a correctness/significance concern, not circularity. Overall, the derivation chain is self-contained with respect to the central empirical claim; only a minor self-citation appears in the generalization to arbitrary dense passive layers.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The model introduces no fitted physical constants: waveguide couplings and lengths are taken from known lattice designs, and phase values are random. The main hand-chosen element is the statistical acceptance threshold (mean ± 1 sigma) that defines 'white noise' and therefore determines whether M=2 is considered sufficient. The architecture also imports the density criterion from the authors' prior work as a domain assumption, and the claim that U is a random matrix is assumed rather than demonstrated.

free parameters (1)
  • White-noise acceptance thresholds = mean ± 1 standard deviation of 1000-sample reference ensembles
    Section 2C defines a signal as white noise when its truncated autocorrelation mean and standard deviation, and Shannon entropy, fall within the shaded bands computed as mean ± 1σ over 1000 normal or Rayleigh reference signals. These thresholds are chosen by hand, and the assessment that M=2 is sufficient depends directly on this choice.
assumptions (4)
  • standard math Coupled-mode theory with a tridiagonal coupling matrix H accurately models the unitary mixing layer F(alpha)=exp(i alpha H).
    Invoked in Section 2A and Methods 4A to justify the waveguide array model.
  • domain assumption The density (Goldilocks) criterion for passive layers guarantees sufficient mixing for interlaced unitary architectures.
    Imported from the authors' prior work (ref 26, Zelaya et al., Sci. Reports 2024) and used to select F; the criterion is qualitative and not re-derived here.
  • domain assumption Entropy and truncated autocorrelation falling inside the reference ensemble bands are sufficient to classify a signal as white noise.
    Defined in Section 2C; used to judge numerical and experimental outputs, and the sufficiency of this classification is assumed without a formal statistical test.
  • domain assumption Random phases drawn uniformly from (-pi, pi] produce a random unitary matrix U whose action whitens any sparse input when M=2.
    Used throughout the simulations; the distribution of U is not characterized, so the 'random matrix' claim rests on this assumption.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Integrated Photonic Programmable Random Matrix Generator with Minimal Active Components." pith.science (2026). https://pith.science/paper/VF2NTBL6

@misc{pith2026250108953,
  author       = {Pith},
  title        = {Pith review of: Integrated Photonic Programmable Random Matrix Generator with Minimal Active Components},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VF2NTBL6}},
  note         = {Machine review of arXiv:2501.08953}
}
read the original abstract

Random matrices are fundamental in photonic computing because of their ability to model and enhance complex light interactions and signal processing capabilities. In manipulating classical light, random operations are utilized for random projections and dimensionality reduction, which are important for analog signal processing, computing, and imaging. In quantum information processing, random unitary operations are essential to boson sampling algorithms for multiphoton states in linear photonic circuits. In photonic circuits, random operations are realized through disordered structures resulting in fixed unitary operations or through large meshes of interferometers and reconfigurable phase shifters, which require a large number of phase shifters. In this article, we introduce a compact photonic circuit for generating random matrices by utilizing programmable phase modulation layers interlaced with a fixed mixing operator. We show that using only two random phase layers is sufficient for producing output optical signals with a white-noise profile, even for highly sparse input optical signals. We experimentally demonstrate these results using a silicon photonics circuit with tunable thermal phase shifters and utilize waveguide lattices as mixing layers. The proposed circuit offers a practical method for generating random matrices for photonic information processing and for applications in data encryption.

Figures

Figures reproduced from arXiv: 2501.08953 by the authors.

Figure 1
Figure 1. Experimental setup and PIC design. a PIC interlaced structure for M = 2 layers. The input optical signal x is fed into the PIC, and the randomization phases are programmed by the SMU controller. The processed optical output signal is z. For complete￾ness, the fully packaged fabricated chip (b), a microscope image of the photonic circuit area (c), and a SEM capture of the waveg￾uide array section (d) are illustrated.… view at source ↗
Figure 2
Figure 2. Experimental run and data processing. a Sequences of testing random pulsed trains x (p) ∈ R100, for p ∈ {1, . . . , 20}. b The latter are demultiplexed into signals xˆ (p) := R20×5 , which are programmed in the N × 1 switch and injected into the PIC. This produces the randomized demultiplexed signals zˆ (p) . c In this process, two ensembles of random phases are loaded into the phase shifters through the SMU, which … view at source ↗
Figure 3
Figure 3. White-noise criteria. (a) Typical autocorrelation profile for white noise signals. (b) Truncated autocorrelation criterion Xe[·] and (c) Shannon entropy (S) for ensembles of normal (blue-shaded) and uniform (purple-shaded) distributions as a function of the distribution size N. (d) Typical autocorrelation profile (upper panel) and the corresponding finite difference ∆ℓ of the modulus of white noise signals. The corr… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Numerical simulations of the device randomness capabilities. (a) Set of 50 randomized pulsed sample signals x n . (b) The corresponding values for the Shannon entropy (upper panel) and autocorrelation bars (lower panel). The shaded area denotes the region where white n…
Figure 5
Figure 5. Figure 5: Random defects and encryption capabilities. (left panel) Real and imaginary parts of the perturbed DFrFT matrix F(δ) for δ = 0.05, 0.1, 0.2. (Top panel) Testing image W used for encryption with perturbed DFrFT F(δ). (right panel) The corresponding encrypted images. (Bo…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

56 extracted references · 54 canonical work pages

  1. [26]

    Zelaya, M

    K. Zelaya, M. Markowitz, and M.-A. Miri, Sci. Reports 14, 10950 (2024)

    work page 2024

  2. [1]

    Carolan, C

    J. Carolan, C. Harrold, C. Sparrow, E. Martín-López, N. J. Russell, J. W. Silverstone, P . J. Shadbolt, N. Matsuda, M. Oguma, M. Itohet al., Science 349, 711 (2015)

    work page 2015

  3. [2]

    Bogaerts, D

    W. Bogaerts, D. Pérez, J. Capmany, D. A. Miller, J. Poon, D. Englund, F . Morichetti, and A. Melloni, Nature 586, 207 (2020)

    work page 2020

  4. [3]

    Bogaerts and A

    W. Bogaerts and A. Rahim, IEEE J. Sel. Top. Quantum Electron. 26, 1 (2020)

    work page 2020

  5. [4]

    T. K. Paraiso, T. Roger, D. G. Marangon, I. De Marco, M. Sanzaro, R. I. Woodward, J. F . Dynes, Z. Yuan, and A. J. Shields, Nat. Photonics15, 850 (2021)

    work page 2021

  6. [5]

    Y . Shen, N. C. Harris, S. Skirlo, M. Prabhu, T. Baehr-Jones, M. Hochberg, X. Sun, S. Zhao, H. Larochelle, D. Englundet al., Nat. photonics 11, 441 (2017)

    work page 2017

  7. [6]

    X. Lin, Y . Rivenson, N. T. Y ardimci, M. Veli, Y . Luo, M. Jarrahi, and A. Ozcan, Science361, 1004 (2018)

    work page 2018

  8. [7]

    Li, J.-X

    X.-K. Li, J.-X. Ma, X.-Y . Li, J.-J. Hu, C.-Y . Ding, F .-K. Han, X.-M. Guo, X. Tan, and X.-M. Jin, Nat. Commun.15, 1044 (2024)

    work page 2024

Show all 56 references
  1. [8]

    Silicon photonic resonator sensors and devices,

    L. Chrostowski, S. Grist, J. Flueckiger, W. Shi, X. Wang, E. Ouellet, H. Yun, M. Webb, B. Nie, Z. Lianget al., “Silicon photonic resonator sensors and devices,” in Laser Resonators, Microresonators, and Beam Control XIV , , vol. 8236 (SPIE, 2012), pp. 387–402

    2012

  2. [9]

    X. Wang, B. Redding, N. Karl, C. Long, Z. Zhu, J. Skowronek, S. Pang, D. Brady, and R. Sarma, Nat. Commun. 15, 4510 (2024)

    2024

  3. [10]

    N. C. Harris, G. R. Steinbrecher, M. Prabhu, Y . Lahini, J. Mower, D. Bunandar, C. Chen, F . N. Wong, T. Baehr-Jones, M. Hochberget al., Nat. Photonics 11, 447 (2017)

    2017

  4. [11]

    M. Wang, C. Doyle, B. Bell, M. J. Collins, E. Magi, B. J. Eggleton, M. Segev, and A. Blanco-Redondo, Nanophotonics 8, 1327 (2019)

    2019

  5. [12]

    M. Reck, A. Zeilinger, H. J. Bernstein, and P . Bertani, Phys. review letters 73, 58 (1994)

    1994

  6. [13]

    D. A. Miller, Photonics Res. 1, 1 (2013)

    2013

  7. [14]

    D. A. Miller, Optica 2, 747 (2015)

    2015

  8. [15]

    W. R. Clements, P . C. Humphreys, B. J. Metcalf, W. S. Kolthammer, and I. A. Walmsley, Optica3, 1460 (2016)

    2016

  9. [16]

    Hamerly, S

    R. Hamerly, S. Bandyopadhyay, and D. Englund, Phys. Rev. Appl. 18, 024019 (2022)

    2022

  10. [17]

    Bogaerts, D

    W. Bogaerts, D. Pérez, J. Capmany, D. A. Miller, J. Poon, D. Englund, F . Morichetti, and A. Melloni, Nature586, 207 (2020)

    2020

  11. [18]

    T. Dai, A. Ma, J. Mao, Y . Ao, X. Jia, Y . Zheng, C. Zhai, Y . Y ang, Z. Li, B. Tanget al., Nat. Mater. pp. 1–9 (2024)

    2024

  12. [19]

    Zhuang, C

    L. Zhuang, C. G. Roeloffzen, M. Hoekman, K.-J. Boller, and A. J. Lowery, Optica 2, 854 (2015)

    2015

  13. [20]

    Pérez, I

    D. Pérez, I. Gasulla, L. Crudgington, D. J. Thomson, A. Z. Khokhar, K. Li, W. Cao, G. Z. Mashanovich, and J. Capmany, Nat. communications8, 636 (2017)

    2017

  14. [21]

    Pérez-López, A

    D. Pérez-López, A. M. Gutierrez, E. Sánchez, P . DasMahapatra, and J. Capmany, Opt. Express 27, 38071 (2019)

    2019

  15. [22]

    Taguchi, Y

    Y . Taguchi, Y . Wang, R. Tanomura, T. Tanemura, and Y . Ozeki, Phys. Rev. Appl.19, 054002 (2023)

    2023

  16. [23]

    Tanomura, R

    R. Tanomura, R. Tang, S. Ghosh, T. Tanemura, and Y . Nakano, J. Light. Technol.38, 60 (2020)

    2020

  17. [24]

    Universal unitary photonic circuits by interlacing discrete fractional fourier transform and phase modulation,

    M. Markowitz and M.-A. Miri, “Universal unitary photonic circuits by interlacing discrete fractional fourier transform and phase modulation,” ArXiv:2307.07101 [physics.optics], 2023

    2023 arXiv

  18. [25]

    Markowitz, K

    M. Markowitz, K. Zelaya, and M.-A. Miri, Opt. Express 31, 37673 (2023)

    2023

  19. [27]

    V. L. Pastor, J. Lundeen, and F . Marquardt, Opt. Express 29, 38441 (2021)

    2021

  20. [28]

    N. C. Harris, Y . Ma, J. Mower, T. Baehr-Jones, D. Englund, M. Hochberg, and C. Galland, Opt. express 22, 10487 (2014)

    2014

  21. [29]

    S. Liu, J. Feng, Y . Tian, H. Zhao, L. Jin, B. Ouyang, J. Zhu, and J. Guo, Front. Optoelectronics 15, 9 (2022)

    2022

  22. [30]

    C. Ríos, Q. Du, Y . Zhang, C.-C. Popescu, M. Y . Shalaginov, P . Miller, C. Roberts, M. Kang, K. A. Richardson, T. Guet al., PhotoniX 3, 26 (2022)

    2022

  23. [31]

    Gigan, Nat

    S. Gigan, Nat. Phys. 18, 980 (2022)

    2022

  24. [32]

    Refregier and B

    P . Refregier and B. Javidi, Opt. letters 20, 767 (1995)

    1995

  25. [33]

    Frauel, A

    Y . Frauel, A. Castro, T. J. Naughton, and B. Javidi, Opt. Express 15, 10253 (2007)

    2007

  26. [34]

    Audhkhasi and M

    R. Audhkhasi and M. L. Povinelli, Adv. Photonics Res. 5, 2300254 (2024)

    2024

  27. [35]

    Audhkhasi, M

    R. Audhkhasi, M. Zhelyeznyakov, S. Brunton, and A. Majumdar, Appl. Opt. 63, G18 (2024)

    2024

  28. [36]

    Miri, Opt

    M.-A. Miri, Opt. Lett. 46, 4936 (2021)

    2021

  29. [37]

    A. Mafi, J. Ballato, K. W. Koch, and A. Schülzgen, J. Light. Technol. 37, 5652 (2019)

    2019

  30. [38]

    Mafi and J

    A. Mafi and J. Ballato, Front. Phys. 9, 736774 (2021)

    2021

  31. [39]

    Cao and Y

    H. Cao and Y . Eliezer, Appl. Phys. Rev. 9 (2022)

    2022

  32. [40]

    Segev, Y

    M. Segev, Y . Silberberg, and D. N. Christodoulides, Nat. Photonics 7, 197 (2013)

    2013

  33. [41]

    A. F . Abouraddy, G. Di Giuseppe, D. N. Christodoulides, and B. E. Saleh, Phys. Rev. A 86, 040302 (2012)

    2012

  34. [42]

    Dikopoltsev, S

    A. Dikopoltsev, S. Weidemann, M. Kremer, A. Steinfurth, H. H. Sheinfux, A. Szameit, and M. Segev, Sci. Adv. 8, eabn7769 (2022)

    2022

  35. [43]

    N. J. Russell, L. Chakhmakhchyan, J. L. O’Brien, and A. Laing, New journal physics 19, 033007 (2017)

    2017

  36. [44]

    Clementi, F

    M. Clementi, F . A. Sabattoli, M. Borghi, L. Gianini, N. Tagliavacche, H. El Dirani, L. Y oussef, N. Bergamasco, C. Petit-Etienne, E. Pargonet al., Nat. Commun. 14, 176 (2023)

    2023

  37. [45]

    X. Meng, G. Zhang, N. Shi, G. Li, J. Azaña, J. Capmany, J. Y ao, Y . Shen, W. Li, N. Zhuet al., Nat. Commun. 14, 3000 (2023)

    2023

  38. [46]

    Zelaya and M.-A

    K. Zelaya and M.-A. Miri, Photonics Res. 12, 1828 (2024)

    2024

  39. [47]

    Taguchi and Y

    Y . Taguchi and Y . Ozeki, JOSA B41, 1425 (2024)

    2024

  40. [48]

    Y ariv and P

    A. Y ariv and P . Y eh,Photonics: optical electronics in modern communications (Oxford university press, 2007)

    2007

  41. [49]

    Huang, JOSA A 11, 963 (1994)

    W.-P . Huang, JOSA A 11, 963 (1994)

    1994

  42. [50]

    M. L. Cooper and S. Mookherjea, Opt. Express 17, 1583 (2009)

    2009

  43. [51]

    Weimann, A

    S. Weimann, A. Perez-Leija, M. Lebugle, R. Keil, M. Tichy, M. Gräfe, R. Heilmann, S. Nolte, H. Moya-Cessa, G. Weihs, D. N. Christodoulides, and A. Szameit, Nat. Commun. 7, 11027 (2016)

    2016

  44. [52]

    D. N. Christodoulides, F . Lederer, and Y . Silberberg, Nature424, 817 (2003)

    2003

  45. [53]

    Forbes, M

    C. Forbes, M. Evans, N. Hastings, and B. Peacock, Statistical distributions (John Wiley & Sons, 2011)

    2011

  46. [54]

    T. M. Cover, Elements of information theory (John Wiley & Sons, 1999)

    1999

  47. [55]

    Unnikrishnan, J

    G. Unnikrishnan, J. Joseph, and K. Singh, Opt. letters 25, 887 (2000)

    2000

  48. [56]

    Belsley, T

    A. Belsley, T. Pertsch, and F . Setzpfandt, Opt. Express 28, 28792 (2020). Funding. This project was supported by the U.S. Air Force Office of Scientific Research (AFOSR) Young Investigator Program (YIP) Award# FA9550- 22-1-0189, the Defense University Research Instrumentation...

    2020

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.