Pith. sign in

REVIEW 2 major objections 5 minor 39 references

Fully stabilized Er fiber comb at 1 GHz by harmonic modelocking

T0 review · 2 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A fiber laser with many pulses in its cavity can make a fully stabilized 1 GHz frequency comb.

desk verdict A credible new architecture for GHz fiber combs, but the long-term Allan deviation claim needs reconciliation with the supplemental data. read the letter →

arxiv 2507.00233 v1 pith:XB3C7USF submitted 2025-06-30 physics.optics

Kevin F. Lee , Jacob Lampen , Peng Li , Jie Jiang , Martin E. Fermann (IMRA America , Inc.) This is my paper · ORCID
classification physics.optics PACS 42.55.Wd42.62.Eh42.60.Fc
keywords frequencycombharmonicmodelockingfiberlasererbiumGHzrepetitionratecarrier-envelopeoffsetsupermodenoiseopticalAllandeviation
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 fiber frequency comb can run at gigahertz repetition rates with many pulses circulating in the laser cavity at once, without the noise that has historically plagued harmonic modelocking. The authors demonstrate a fully stabilized erbium fiber comb at 1.04 GHz whose out-of-loop beat with a reference comb has a linewidth below 0.1 Hz and whose optical Allan deviation averages down from $10^{-17}$ at one second to $10^{-20}$ in an hour. The point is to separate the repetition-rate-determining cavity from the gain medium, so mature fiber-laser technology can reach high repetition rates that previously required solid-state lasers, external filtering cavities, or microresonators. If correct, the result turns repetition rate into a freely selectable design parameter and opens a route to lower timing jitter through the filtering intrinsic to the two-cavity design.

What carries the argument

The key mechanism is the two-cavity design: a short free-space subcavity, roughly 15 cm long and setting the approximately 1 GHz repetition rate, coupled by a beamsplitter to a much longer nonlinear amplifying loop mirror (NALM) fiber gain cavity with a fundamental rate near 104 MHz. Harmonic modelocking means the gain-cavity round-trip time is an integer multiple (here 10) of the subcavity round-trip time, so returning amplified pulses reinforce the subcavity pulse at the beamsplitter. The interference at the beamsplitter is a highly selective filter: light that does not match the subcavity in wavelength, phase, or timing is rejected rather than left in the cavity, which is what suppresses supermode noise and, in the numerical model, reduces quantum-noise-driven timing jitter. The subcavity acts as a low-noise pulse reservoir, and the whole arrangement is described as pulsed self-injection locking, coupling the laser field to a delayed version of itself.

What would settle it

Measure the out-of-loop beat between the GHz comb and a reference comb with a phase-noise analyzer spanning offset frequencies around the gain-cavity free spectral range, about 104 MHz; clean harmonic modelocking predicts no supermode sidebands there, while any residual supermode noise would appear as peaks at those offsets.

Watch

Extended reading notes

Core claim

The central discovery is that harmonic modelocking, in which several pulses circulate in the gain cavity, can be made to behave like a single-pulse comb by adding a short subcavity that governs the repetition rate and interferometrically filters the returning amplified pulses. The subcavity holds a single pulse, so adjacent output pulses are coherent; the longer NALM fiber cavity provides gain and modelocking without its length constraining the repetition rate. The beamsplitter that couples the two cavities acts as a highly selective filter that rejects amplified light not matching the subcavity in wavelength, phase, or timing, which suppresses supermode noise. The paper shows that with the CEO frequency locked via pump current and fiber stretcher and the optical beat locked via subcavity actuators, the resulting comb is fully stabilized and its out-of-loop comparison with a conventional 200 MHz fiber comb shows a beat linewidth below 0.1 Hz and an Allan deviation reaching $10^{-20}$ in an hour. The supplement's numerical model indicates that the same filtering should reduce fundamental timing jitter in proportion to the beamsplitter reflectivity, with a tenfold jitter reduction at 10 percent reflectivity.

Load-bearing premise

The design assumes that multiple pulses circulating in the gain cavity can be treated as independent, with each pulse experiencing the same nonlinear effects it would if it were the only pulse, provided the pulse energy is held constant; if pulse interactions or gain competition instead couple the pulses, harmonic modelocking would reintroduce supermode noise and the claimed comb quality would not hold.

Editorial extensions

If this is right

  • Fiber comb technology can operate at repetition rates from the fundamental up to at least the 12th harmonic, roughly 1.2 GHz, simply by changing the subcavity length, giving more than an order of magnitude of range with a single laser.
  • A dual comb for spectroscopy could be made by polarization duplexing in a shared subcavity with birefringence, providing an ultrastable repetition-rate difference.
  • The filtering intrinsic to the beamsplitter should reduce fundamental timing jitter compared to a standard modelocked laser, with the reduction growing as the beamsplitter reflectivity is lowered.
  • An empty V-shaped subcavity can serve as a passively stable internal reference, potentially bringing some of the stability of high-finesse reference cavities directly into the oscillator.
  • The approach extends to microresonator gain loops, where harmonic modelocking could remove the continuous-wave background and improve pumping efficiency.

Reading between the lines

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

  • The paper's separation of optical cavity and gain medium suggests a general design principle: any short optical resonator can be turned into the repetition-rate-defining element of a modelocked laser without paying the power penalty of external cavity locking.
  • Because the interference filter rejects noise at a rejection port, the design might combine with balanced detection at that port to provide an in-loop error signal with a reduced noise floor, potentially improving the achievable comb linewidth beyond what was demonstrated.
  • The claim that timing jitter scales with beamsplitter reflectivity is testable at lower reflectivities; a 10 percent reflectivity run should show a clear jitter reduction compared to a standard NALM comb, confirming the noise-filtering picture.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 5 minor

Summary. This paper reports a harmonically modelocked Er-fiber comb in which a short free-space subcavity sets a 1.04 GHz repetition rate while a long NALM fiber gain cavity supplies gain and modelocking, so that multiple pulses circulate in the gain cavity. The authors fully stabilize the CEO and repetition-rate degrees of freedom and characterize the laser with spectra, RIN, in-loop and out-of-loop optical beats (0.23, 0.48, and 0.36 rad integrated phase noise; out-of-loop linewidth below 0.1 Hz), and an optical Allan deviation reported to average from 1e-17 at 1 s to 1e-20 at 1 h. They also describe a numerical model predicting reduced timing jitter at low beamsplitter reflectivity and demonstrate two design variants, a fiberized subcavity and an empty V-shaped subcavity.

Significance. If the long-term stability result survives scrutiny, this design extends mature Er-fiber comb technology to GHz repetition rates while retaining conventional comb performance and offers repetition-rate tuning over more than an order of magnitude. The separation of the repetition-rate-defining subcavity from the gain medium is an elegant form of harmonic modelocking, and the beamsplitter rejection port is a plausible mechanism for supermode suppression. The numerical model is clearly presented and makes falsifiable predictions, and the experimental characterization includes out-of-loop verification rather than only in-loop locks. The main weaknesses are that the headline Allan deviation is not supported by the raw data shown, and the supermode and jitter benefits are not quantified experimentally.

major comments (2)
  1. [Section 4, Fig. 3, and Supplement S4] The headline long-term stability claim is not supported by the presented data. The 3-hour out-of-loop record in Fig. S4 stays within roughly 15 mHz of its mean, and the text states that noncommon-fiber phase drifts were reduced to 'below 1 Hz'; at 191 THz the claimed 1e-20 at 1 h corresponds to about 2 microhertz, which is orders of magnitude below both numbers. Please provide the raw counter data, the exact Allan deviation computation (including any detrending, dead-time, or averaging procedure), and an explicit measurement-noise floor for the out-of-loop path, separating comb stability from environmental phase shifts.
  2. [Section 3 and Supplement S2] Clean harmonic modelocking is inferred from RF spectra at 30 kHz resolution, but no quantitative supermode-suppression measurement is reported, and the predicted timing-jitter reduction at low beamsplitter reflectivity is not experimentally verified. Since the paper claims noise filtering and potential jitter improvements as advantages, please add a quantitative supermode characterization (e.g., sideband-to-carrier ratio or a phase-noise spectrum spanning the gain-cavity FSR) and explicitly distinguish the simulated jitter prediction from measured performance.
minor comments (5)
  1. [Fig. 1c] The RIN peak near 300 Hz is left unexplained; a sentence on its origin would help the reader assess whether it is a fundamental or technical noise source.
  2. [Section 2] The statement that the CEO frequency is highly dependent on the relative cavity matching would be more convincing with a quantitative plot of CEO frequency versus cavity-length offset; the current actuator-assignment explanation is qualitative.
  3. [Supplement S2] The model compresses the temporal window to 0.1-1 ns; please justify why this compression does not alter the simulated nonlinear interaction and state the numerical grid and round-trip counts used.
  4. [Fig. 3 and Fig. 2 captions] Please specify the integration bandwidth for the phase-noise values and add confidence intervals or error bars to the Allan deviation, particularly at the longest averaging times.
  5. [General] The paper does not include a data-availability statement; providing the raw counter records and analysis script would materially strengthen reproducibility.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: the jitter-reduction model is a forward prediction and the out-of-loop comparison uses an external reference comb; minor self-citations are not load-bearing.

full rationale

The central claim is an experimental demonstration: a harmonically modelocked GHz fiber comb is fully stabilized and compared out-of-loop against a conventional 200 MHz reference comb. The reference comb, its Rb-stabilized repetition rate, and the 1556 nm/1565 nm CW lasers are external to the GHz comb under test, so the out-of-loop beat, linewidth, and Allan deviation are not defined in terms of the GHz comb's own fitted parameters. The numerical model in Supplement S1-S2 is also not used to fit the measured comb performance: parameters such as pigtail length, loop length, gamma, and beta2 are chosen to 'approximately match the experimental configuration', but the predicted quantity (fundamental timing jitter reduction versus beamsplitter reflectivity, Fig. S2) is a forward simulation, not a retrodiction of the measured Fig. 3 Allan deviation or the Fig. 2 beat linewidths. No equation in the paper sets the measured stability equal to a model input. The statement that 'the nonlinear effects on each pulse match those for a single circulating pulse, so multiple pulses can exist independently within the laser' is explicitly an assumption ('Assuming enough pumping to have the same pulse energy at any repetition rate'), with an acknowledged failure mode (supermode noise), rather than a disguised fit. The self-citations to [22], [23], and [27] are not load-bearing: [23] is used mainly as a comparison benchmark ('the best NALM oscillators with better actuators can be well below 0.1 rad'), [22] is the original NALM reference, and [27] supports the general practice of locking cavities; none of these are invoked to forbid alternatives or to define the present result. The paper does contain an internal-consistency concern raised by the skeptic: the paper states 'At these high stability levels, the results may be measurement limited' and Fig. S4 shows a ~15 mHz range over 3 hours, which is hard to reconcile with the 1e-20-at-1-hour Allan deviation in Fig. 3 without more analysis details. That is a data-analysis and verifiability issue, not a circularity issue: it concerns whether the long-term stability claim is correctly computed or independently reproducible, not whether the claim reduces by construction to its inputs. No raw counter data or analysis code are provided, so the long-term claim is difficult to audit externally, but this does not make the derivation circular.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The central experimental claim rests on the domain assumptions of independent multiple pulsing and interference-based filtering. The only explicit free parameter in the model is the gain saturation pair, which is not specified, and no new physical entities are introduced.

free parameters (1)
  • gain saturation parameters g0 and Esat = unspecified
    Used in the Supplement's NALM simulation to estimate timing jitter; values are not provided, limiting reproducibility of the jitter prediction.
assumptions (3)
  • domain assumption Multiple pulses in the gain cavity can be treated as independent, with nonlinear effects matching those of a single pulse when pulse energy is maintained.
    Stated in the Introduction as the basis for harmonic modelocking; underpins the entire design.
  • domain assumption The interference condition at the beamsplitter provides a highly selective filter that suppresses supermode noise.
    Central to the claim of clean harmonic modelocking; supported indirectly by RF spectra but not directly measured.
  • domain assumption The round-trip time of the gain cavity is an integer multiple of the subcavity round-trip time (N=10 in the main experiment).
    Required for the subcavity pulse to meet the returning amplified pulse; the lock must maintain this relation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Fully stabilized Er fiber comb at 1 GHz by harmonic modelocking." pith.science (2026). https://pith.science/paper/XB3C7USF

@misc{pith2026250700233,
  author       = {Pith},
  title        = {Pith review of: Fully stabilized Er fiber comb at 1 GHz by harmonic modelocking},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XB3C7USF}},
  note         = {Machine review of arXiv:2507.00233}
}
read the original abstract

Modelocked frequency comb lasers have always operated with a single pulse circulating in the laser cavity. This meant that each laser technology had an associated limit on pulse repetition rate. Achieving higher rates required different technology, for example exchanging optical fiber for solid state gain media. However, this is a perceived, rather than a fundamental limit. We demonstrate a new fiber laser design with multiple pulses circulating in the fiber gain cavity with the same high precision as a conventional fiber frequency comb. This has the immediate benefit of bringing mature fiber technology to higher repetition rate frequency combs. More generally, it adds great design freedom to laser engineering, where the laser can be separated into an optical cavity and a gain medium that are combined using standard frequency comb techniques. Fundamental frequency comb performance improvements may even be possible from the filtering intrinsic to our design, or by incorporating high stability cavities directly into the laser itself.

Figures

Figures reproduced from arXiv: 2507.00233 by the authors.

Figure 1
Figure 1. GHz laser showing basic characteristics quite similar to an ordinary fiber comb. (a) Drawing of harmonically modelocked GHz fiber laser design using a short freespace subcavity (right, blue shade) to increase repetition rate of a conventional NALM fiber oscillator (left, orange shade). (b) Spectrum of laser output at 1 GHz repetition rate supporting femtosecond pulses. (c) Relative intensity noise of 1 GHz comb, wit… view at source ↗
Figure 2
Figure 2. (a) locking and interference arrangement for harmonically modelocked GHz comb including a standard fiber comb for comparison. Both combs have stabilized CEO and are locked to a 1556 nm optical reference which is stabilized against the reference comb repetition rate. The distance between comb lines from each comb is measured by mixing the beat signals of each comb with a 1565 nm laser which is locked to a reference c… view at source ↗
Figure 3
Figure 3. Allan deviation of frequency difference between a comb line from the GHz comb and a reference frequency comb. The deviation decreases steadily, indicating good long term stability. nm laser to the reference comb. A high resolution RF spectrum of the mixed beat note is shown in Fig. 2b. The linewidth is below 0.1 Hz, and the integrated phase noise from 3.5 MHz is 0.36 rad, similar to the phase noise of the locked bea… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Variations of our harmonic modelocking design with different subcavity types. (a) Fiberizing the subcavity (top, blue shade) to approach an all fiberized GHz comb, but the mirrors are currently still freespace. The gain cavity is drawn simplified and shaded orange. (b)…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 37 canonical work pages

  1. [1]

    Chang, S

    L. Chang, S. Liu, and J. E. Bowers, Integrated optical frequency comb technologies, Nature Photonics 16, 95--108 (2022)

    work page 2022

  2. [2]

    Fortier and E

    T. Fortier and E. Baumann, 20 years of developments in optical frequency comb technology and applications, Communications Physics 2, 153 (2019)

    2019

  3. [3]

    Wilken, G

    T. Wilken, G. L. Curto, R. A. Probst, et al., A spectrograph for exoplanet observations calibrated at the centimetre-per-second level, Nature 485, 611--614 (2012)

    work page 2012

  4. [4]

    Corcoran, M

    B. Corcoran, M. Tan, X. Xu, et al., Ultra-dense optical data transmission over standard fibre with a single chip source, Nature communications 11, 2568 (2020)

    work page 2020

  5. [5]

    Hartl, H

    I. Hartl, H. McKay , R. Thapa, et al., GHz Yb -femtosecond-fiber laser frequency comb, in 2009 Conference on Lasers and Electro-Optics and 2009 Conference on Quantum electronics and Laser Science Conference, (IEEE, 2009), pp. 1--2

    work page 2009

  6. [6]

    Jiang, A

    T. Jiang, A. Wang, G. Wang, et al., Tapered photonic crystal fiber for simplified Yb : fiber laser frequency comb with low pulse energy and robust f ceo singals, Optics Express 22, 1835--1841 (2014)

    work page 2014

  7. [7]

    D. M. Lesko, A. J. Lind, N. Hoghooghi, et al., Fully phase-stabilized 1 GHz turnkey frequency comb at 1.56 m, OSA continuum 3, 2070--2077 (2020)

    work page 2020

  8. [8]

    M \"u ller, M

    M. M \"u ller, M. Hamrouni, K. N. Komagata, et al., Powerful 1- m 1- GHz optical frequency comb, Optics Express 31, 44823--44831 (2023)

    work page 2023

Show all 39 references
  1. [9]

    Zhang, C

    J. Zhang, C. Wang, C. Denney, et al., Ultrabroadband integrated electro-optic frequency comb in lithium tantalate, Nature 637, 1096--1103 (2025)

    2025

  2. [10]

    A. L. Gaeta, M. Lipson, and T. J. Kippenberg, Photonic-chip-based frequency combs, nature photonics 13, 158--169 (2019)

    2019

  3. [11]

    Becker, D

    M. Becker, D. Kuizenga, and A. Siegman, Harmonic mode locking of the Nd: YAG laser, IEEE Journal of Quantum Electronics 8, 687--693 (1972)

    1972

  4. [12]

    Kafka, T

    J. Kafka, T. Baer, and D. W. Hall, Mode-locked erbium-doped fiber laser with soliton pulse shaping, Optics letters 14, 1269--1271 (1989)

    1989

  5. [13]

    Takada and H

    A. Takada and H. Miyazawa, 30 GHz picosecond pulse generation from actively mode-locked erbium-doped fibre laser, Electronics Letters 26, 216--217 (1990)

    1990

  6. [14]

    G. T. Harvey and L. F. Mollenauer, Harmonically mode-locked fiber ring laser with an internal fabry--perot stabilizer for soliton transmission, Optics letters 18, 107--109 (1993)

    1993

  7. [15]

    S. Gee, F. Quinlan, S. Ozharar, and P. Delfyett, Simultaneous optical comb frequency stabilization and super-mode noise suppression of harmonically mode-locked semiconductor ring laser using an intracavity etalon, IEEE Photonics Technology Letters 17, 199--201 (2004)

    2004

  8. [16]

    Hasegawa, Phase structure of harmonically mode-locked optical frequency combs, Optics Express 32, 8092--8100 (2024)

    T. Hasegawa, Phase structure of harmonically mode-locked optical frequency combs, Optics Express 32, 8092--8100 (2024)

    2024

  9. [17]

    Korobko, V

    D. Korobko, V. Ribenek, D. Stoliarov, et al., Resonantly induced mitigation of supermode noise in a harmonically mode-locked fiber laser: revealing the underlying mechanisms, Optics Express 30, 17243--17258 (2022)

    2022

  10. [18]

    Rowley, P.-H

    M. Rowley, P.-H. Hanzard, A. Cutrona, et al., Self-emergence of robust solitons in a microcavity, Nature 608, 303--309 (2022)

    2022

  11. [19]

    X. Cao, J. Zhou, Z. Cheng, et al., GHz figure-9 er-doped optical frequency comb based on nested fiber ring resonators, Laser & Photonics Reviews 17, 2300537 (2023)

    2023

  12. [20]

    Z. Ding, Z. Huang, Y. Chen, et al., All-fiber ultrafast laser generating gigahertz-rate pulses based on a hybrid plasmonic microfiber resonator, Advanced Photonics 2, 026002--026002 (2020)

    2020

  13. [21]

    Cutrona, M

    A. Cutrona, M. Rowley, A. Bendahmane, et al., Stability of laser cavity-solitons for metrological applications, Applied Physics Letters 122 (2023)

    2023

  14. [22]

    M. E. Fermann, F. Haberl, M. Hofer, and H. Hochreiter, Nonlinear amplifying loop mirror, Opt. Lett. 15, 752--754 (1990)

    1990

  15. [23]

    N. Kuse, J. Jiang, C.-C. Lee, et al., All polarization-maintaining Er fiber-based optical frequency combs with nonlinear amplifying loop mirror, Optics express 24, 3095--3102 (2016)

    2016

  16. [24]

    Dahmani, L

    B. Dahmani, L. Hollberg, and R. Drullinger, Frequency stabilization of semiconductor lasers by resonant optical feedback, Optics letters 12, 876--878 (1987)

    1987

  17. [25]

    Maleki, Novel lasers: Whispering-gallery-mode resonators create ultranarrow-linewidth semiconductor lasers, Laser Focus World (2014)

    L. Maleki, Novel lasers: Whispering-gallery-mode resonators create ultranarrow-linewidth semiconductor lasers, Laser Focus World (2014)

    2014

  18. [26]

    Oelker, R

    E. Oelker, R. Hutson, C. Kennedy, et al., Demonstration of 4.8 10^ -17 stability at 1 s for two independent optical clocks, Nature Photonics 13, 714--719 (2019)

    2019

  19. [27]

    K. F. Lee, G. Kowzan, C. Lee, et al., Frequency combs for cavity cascades: Opo combs and graphene-coupled cavities, Journal of Physics B: Atomic, Molecular and Optical Physics 50, 014003 (2016)

    2016

  20. [28]

    L. C. Sinclair, J.-D. Desch \^e nes, L. Sonderhouse, et al., Invited article: A compact optically coherent fiber frequency comb, Review of scientific instruments 86 (2015)

    2015

  21. [29]

    Burkart, D

    J. Burkart, D. Romanini, and S. Kassi, Optical feedback stabilized laser tuned by single-sideband modulation, Optics letters 38, 2062--2064 (2013)

    2013

  22. [30]

    J. Liu, H. Tian, Q.-X. Ji, et al., Phase stabilization of a microcomb with integrated piezoelectric actuators, CLEO 2025 p. AA123.7 (2025)

    2025

  23. [31]

    J. Li, C. Bao, Q.-X. Ji, et al., Efficiency of pulse pumped soliton microcombs, Optica 9, 231--239 (2022)

    2022

  24. [32]

    Zhang, C

    X. Zhang, C. Wang, Z. Cheng, et al., Advances in resonator-based kerr frequency combs with high conversion efficiencies, npj Nanophotonics 1, 26 (2024)

    2024

  25. [33]

    Z. Wang, Q. Nie, H. Sun, et al., Cavity-enhanced photoacoustic dual-comb spectroscopy, Light: Science & Applications 13, 11 (2024)

    2024

  26. [34]

    Allan and R

    E. Allan and R. A. McCracken, Low-repetition-rate near-infrared femtosecond pulse generation in a harmonically pumped idler-resonant optical parametric oscillator, Optics Continuum 3, 1509--1515 (2024)

    2024

  27. [35]

    M. C. Gather and S. H. Yun, Single-cell biological lasers, Nature Photonics 5, 406--410 (2011)

    2011

  28. [36]

    Paschotta, Noise of mode-locked lasers (Part I) : numerical model, Applied Physics B 79, 153--162 (2004)

    R. Paschotta, Noise of mode-locked lasers (Part I) : numerical model, Applied Physics B 79, 153--162 (2004)

    2004

  29. [37]

    Paschotta, A

    R. Paschotta, A. Schlatter, S. Zeller, et al., Optical phase noise and carrier-envelope offset noise of mode-locked lasers, Applied Physics B 82, 265--273 (2006)

    2006

  30. [38]

    Del'Haye, A

    P. Del'Haye, A. Coillet, T. Fortier, et al., Phase-coherent microwave-to-optical link with a self-referenced microcomb, Nature Photonics 10, 516--520 (2016)

    2016

  31. [39]

    R. M. Gray, R. Sekine, M. Shen, et al., Two-optical-cycle pulses from nanophotonic two-color soliton compression, arXiv preprint arXiv:2501.15381 (2025)

    2025 arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.