Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

A Bayesian perspective on single-shot laser characterization

T0 review · 4 major / 5 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read Single-shot capability in laser metrology is a threshold, set when measurement precision falls below twice the intrinsic shot-to-shot fluctuation — not a built-in property of the device.

desk verdict A useful conceptual framework for when a measurement truly resolves single shots, with a plausible but under-validated demonstration; worth reviewing, not desk-rejecting. read the letter →

arxiv 2502.03100 v1 pith:6BP2EDHE submitted 2025-02-05 physics.optics cs.LGphysics.ins-det

J. Esslinger , N. Weisse , C. Eberle , J. Schroeder , S. Howard , P. Norreys , S. Karsch , A. Döpp This is my paper · ORCID
classification physics.opticscs.LGphysics.ins-det
keywords Bayesianinferencesingle-shotmeasurementspatio-temporalcouplingslasermetrologyuncertaintyquantificationpulsefronttiltcurvaturestate-resolvingregime
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

Laser metrology usually labels a device 'single-shot' or 'multi-shot' by how many pulses it consumes. This paper argues that label is wrong: the real question is whether a single measurement can resolve the instantaneous state of the laser, and that is decided by the ratio of measurement noise to the laser's intrinsic shot-to-shot variance. When the measurement variance exceeds twice the stochastic variance, $\sigma_{\mathrm{meas}}^2 > 2\sigma_{\mathrm{stoch}}^2$, measurements can only describe the statistical distribution; below that threshold, individual shots become distinguishable. The authors build a Bayesian estimator around this threshold, apply it to spatio-temporal couplings at the ATLAS-3000 petawatt laser with a mosaic-filter Shack-Hartmann device, and obtain the first quantitative uncertainty bounds on pulse front tilt and curvature, shrinking posterior uncertainty by up to 60% relative to pseudo-inverse least squares.

What carries the argument

The load-bearing mechanism is the recursive Bayesian update for a Gaussian state-space model, whose self-conjugacy yields an analytic posterior. The state of each Zernike–Taylor coefficient is updated by $\mu_{k|k} = (1-\gamma_k)\mu_{k|k-1} + \gamma_k y_k$ with $\gamma_k = \sigma^2_{k|k-1}/(\sigma^2_{k|k-1} + \sigma^2_{\mathrm{meas}})$. From this update the paper derives the asymptotic variance $\sigma_{\infty}^2$ and the threshold $\sigma_{\mathrm{meas}}^2 = 2\sigma_{\mathrm{stoch}}^2$; the threshold is what separates the distribution-centric from the state-resolving regime. Around this core, the framework couples a local linear model (Holt's exponential smoothing) that tracks the deterministic drift $f(t_k)$, with a maximum-likelihood estimator of the intrinsic stochasticity whose residual variance combines prediction and measurement uncertainty, $\sigma^2_{\Delta\mu} = \sigma^2_{\mathrm{pred}} + \sigma^2_{\mathrm{meas}} \approx \sigma^2_{\mathrm{posterior}} + \sigma^2_{\mathrm{stoch}} + \sigma^2_{\mathrm{meas}}$. In the vectorial case, the same update becomes matrix equations with transfer matrix $T$ and pseudo-inverse $T^+$, giving posterior covariances for the retrieved coefficients.

What would settle it

A controlled test would fix a single mode (e.g., pulse front tilt) and take two measurement series with deliberately different measurement noise, one with $\sigma_{\mathrm{meas}}^2 = 0.5\,\sigma_{\mathrm{stoch}}^2$ and one with $\sigma_{\mathrm{meas}}^2 = 3\,\sigma_{\mathrm{stoch}}^2$; the theory predicts the asymptotic posterior variance falls below $\sigma_{\mathrm{stoch}}^2$ only in the first case. If the state-resolving regime appears in the high-noise setting, or fails to appear in the low-noise setting, the threshold $\sigma_{\mathrm{meas}}^2 = 2\sigma_{\mathrm{stoch}}^2$ is wrong.

Watch

Extended reading notes

Core claim

The central claim is that 'single-shot' capability is an emergent property of the ratio between measurement precision and intrinsic stochasticity, not a fixed attribute of the apparatus. For a Gaussian state-space model where each shot is $x_k = f(t_k) + \varepsilon_k$ with $\varepsilon_k \sim \mathcal{N}(0, \sigma_{\mathrm{stoch}}^2)$ and each measurement is $y_k = x_k + \epsilon_k$ with $\epsilon_k \sim \mathcal{N}(0, \sigma_{\mathrm{meas}}^2)$, the Bayesian posterior variance has an asymptotic limit $\sigma_{\infty}^2 = \frac{\sqrt{1 + 4(\sigma_{\mathrm{meas}}^2/\sigma_{\mathrm{stoch}}^2)} - 1}{2}\,\sigma_{\mathrm{stoch}}^2$. When $\sigma_{\mathrm{meas}}^2 > 2\sigma_{\mathrm{stoch}}^2$ this limit stays above $\sigma_{\mathrm{stoch}}^2$, so repeated measurements only characterize the average system (distribution-centric regime); when $\sigma_{\mathrm{meas}}^2 < 2\sigma_{\mathrm{stoch}}^2$ the limit drops below $\sigma_{\mathrm{stoch}}^2$, meaning a single measurement can resolve the instantaneous state (state-resolving regime). The paper further shows that the threshold coincides with the Bayesian update weight $\gamma = 1/2$ and with a Nyquist-frequency attenuation of about 67%. Demonstrated at the ATLAS-3000 petawatt laser with a single-shot mosaic-filter Shack-Hartmann sensor, this yields per-mode uncertainty bounds on spatio-temporal couplings—the first such quantitative error bars—and the posterior uncertainty on pulse front tilt and curvature is 54–60% smaller than the raw measurement uncertainty.

Load-bearing premise

The framework assumes that after removing the local linear trend, the remaining shot-to-shot fluctuations are independent and Gaussian with a fixed, diagonal covariance, and that this stochastic variance can be estimated reliably from the same data; if the deterministic drift is not captured well by the local linear model, the estimated stochasticity is inflated and the regime classification and uncertainty bounds become biased.

Editorial extensions

If this is right

  • For a given laser mode, the regime boundary is a number, not a label: the same device can be state-resolving for tilt while distribution-centric for linear coma, exactly as observed in the ATLAS-3000 data.
  • With per-shot posterior distributions, spatio-temporal couplings become inputs to closed-loop control: the posterior mean plus its uncertainty can drive beam stabilization, and modes whose fluctuations are stochastic set a floor that no a priori correction can remove.
  • The measured pulse-front-tilt stochasticity corresponds to a 0.3–5.5% loss in focal intensity at ATLAS-3000, giving a quantitative lower bound on peak-intensity fluctuations for an otherwise perfect laser.
  • Because the threshold depends on the ratio, improving the sensor (smaller spot-finding error or better calibration of the model-fit term) can move a mode across the boundary without touching the laser, so the framework doubles as a design rule for diagnostics.
  • The frequency-response analysis implies a tunable trade-off: increasing the update weight improves temporal resolution of shot-to-shot dynamics, while decreasing it suppresses measurement noise; at the threshold the Nyquist frequency is attenuated to roughly 67%, which can serve as an operating-point guideline.

Reading between the lines

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

  • The same Gaussian update and threshold argument should apply to any metrology scheme where a parameter carries both measurement noise and process noise—e.g., beam-position monitors, wavefront sensors in adaptive optics, or even particle-beam diagnostics—so the definition of 'single-shot' could be exported beyond laser science.
  • Since the paper diagonalizes the stochasticity matrix for simplicity and calls the resulting uncertainties upper bounds, a natural extension is a full covariance or Gaussian-process model over time; this could lower the bounds further and reveal correlations between coupled Zernike modes.
  • The separating point in the spectra of Fig. 7 (around 8 minutes for pulse front tilt, 20–40 minutes for linear coma) suggests that the framework could be used online to detect when unresolved dynamics are masquerading as stochasticity, a use the paper does not explicitly develop.
  • The calibrated measurement-noise decomposition (focal-spot-position error plus a model-fit term of 0.6 pixels) is itself a new component for STC diagnostics; applying the same calibration to other sensors would let facilities compare their 'single-shotness' on a common scale.
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

4 major / 5 minor

Summary. The paper introduces a Bayesian state-space framework for laser metrology in which each measured parameter is decomposed into a deterministic trend f(t_k) and an independent Gaussian stochastic component with variance sigma_stoch^2, while each measurement adds Gaussian noise sigma_meas^2. Section II derives the scalar Kalman-type update, the steady-state posterior variance in Eq. (8), and the threshold sigma_meas^2 = 2 sigma_stoch^2 separating a distribution-centric from a state-resolving regime. It also presents a frequency-response interpretation and a residual-based maximum-likelihood estimate of sigma_stoch using a local linear (Holt) model. Section III describes a new 'single-shot FALCON' device built from a mosaic 3x3 filter array in front of a Shack-Hartmann sensor, applies the framework at the ATLAS-3000 petawatt laser, and reports posterior-variance reductions of 47--60% compared with pseudo-inverse least squares for pulse-front tilt, curvature, and related modes. The paper claims this provides the first quantitative uncertainty bounds for spatio-temporal coupling measurements.

Significance. If the validation gaps are closed, this is a useful conceptual contribution: it replaces the binary single-shot/multi-shot distinction with a quantitative criterion based on the ratio sigma_meas^2 / sigma_stoch^2, and it produces per-shot posterior bounds for spatio-temporal couplings, which are currently missing from the STC literature. The analytic derivation leading to Eq. (8) is compact, transparent, and internally consistent, and the mosaic-filter device is simple and inexpensive. The paper also deserves credit for an explicit calibration of the measurement-noise covariance, something that is often omitted in STC diagnostics. However, the empirical support for the central quantitative claims is currently limited to self-consistency checks: the intrinsic stochasticity is estimated from the residuals of the same model used for prediction, and the reported uncertainty reduction follows mathematically from the Bayesian update rule. These issues must be addressed before the quantitative regime classification and the claimed first uncertainty bounds are fully established.

major comments (4)
  1. [Section II D and Section II E (Eqs. (11), (16))] The estimate of sigma_stoch is obtained from the residuals of the same local-linear model that is used to form the predictions. Equation (11) defines the residual variance as sigma_posterior^2 + sigma_stoch^2 + sigma_meas^2, but any deterministic trend not captured by the N=20 local-linear fit (e.g., nonlinear thermal drift or slow oscillations) enters the residual and is absorbed into sigma_stoch. The text itself acknowledges this in Section II D: 'errors in estimating f(tk) will be absorbed into our estimate of epsilon_k.' Because the regime classification of Section II B and the posterior bounds of Section III E both depend directly on sigma_stoch, this circularity is load-bearing. The manuscript provides only a self-consistency check (histograms matching Eq. (11)), not an independent validation of the trend--stochastic decomposition. I recommend adding a synthetic-data test with known f(t) and sigma_stoch, plus an independent estimate of sigma_stoch, for example from a second faster diagnostic or from held-out windows, before the reported regime assignment and the uncertainty bounds can be trusted.
  2. [Section III E (Figs. 5 and 9)] The reported 54--60% 'uncertainty reduction' is a direct mathematical consequence of Eqs. (5)--(6), not an empirical validation of accuracy. Any informative prior produces a posterior variance smaller than the measurement variance; comparing the posterior variance with the pseudo-inverse's measurement-noise variance quantifies only the amount of prior shrinkage. The histograms in Fig. 5 show that residuals are consistent with Eq. (11), but this is a self-consistency check of the Gaussian assumptions, not evidence that the posterior intervals are calibrated. To support the claim of 'first quantitative uncertainty bounds,' the authors should validate coverage on data with known parameter values, for example by injecting controlled STC changes, or compare posterior predictive intervals against independent measurements.
  3. [Section II D and Section II F] The assertion that assuming a diagonal intrinsic-stochasticity matrix makes the reconstructed uncertainties 'upper bound values' is not generally valid. Even if mode correlations are neglected, model misspecification in the trend estimate can bias sigma_stoch downward: in the state-resolving regime, each measurement contains a large fraction of the instantaneous fluctuation epsilon_k, and the local-linear fit of Eqs. (14)--(15) can partly absorb that fluctuation into the trend, yielding an overconfident posterior. The opposite bias, inflation of sigma_stoch by unmodeled drift, is also possible. The manuscript argues only one direction of the bias; a formal sensitivity analysis or a simulation with a known trend is needed to establish that the reported bounds are actually conservative.
  4. [Section III C (Eqs. (20)--(22))] The passage from the scalar update to the vectorial update assumes that T^+ d_k is a sufficient statistic for the modal coefficients. If the transfer matrix T in Eq. (19) is rank-deficient or severely ill-conditioned for the 3x3 sub-aperture configuration, the pseudo-inverse can remove or distort information in the null space, and Eq. (21) is not necessarily the full posterior covariance of the original model in Eq. (19). The paper states that the reconstruction becomes 'well-posed' but does not report the rank or condition number of T, nor does it verify that the coefficient-space update is equivalent to the full Bayesian solution under the assumed noise model. This should be demonstrated explicitly, because the quantitative posterior covariance is the main experimental output.
minor comments (5)
  1. [Introduction and Section III E] There are several wording errors: 'as much diverse information as possible within in a single shot' and 'a few- or multi-shot device can also be a reasonable choice' should be corrected.
  2. [Supplemental Material, Fig. 9] The upper-right panel is labeled 'Pulse Front Curvature (PFT)/linear defocus'; PFT is used elsewhere in the paper for pulse front tilt, so this label should be corrected (likely 'PFC') to avoid confusion.
  3. [Figs. 5 and 9] The zoomed panels show the posterior mean and pseudo-inverse mean but do not display uncertainty bands or error bars; showing the posterior uncertainty would make the claimed reduction visible and easier to assess.
  4. [Eq. (24)] The formula for chi_intensity appears to have an unmatched parenthesis; the outer exponent should apply to the entire factor (1 + ...), not just to the numerator term, and the notation should be unambiguous.
  5. [Eq. (12) and Fig. 5] The residual model assumes independent, Gaussian residuals, but the paper does not comment on autocorrelation of the residuals shown in Fig. 5. A lag-correlation plot or Durbin--Watson-type statistic would strengthen the self-consistency check and support the diagonal-covariance assumption.

Circularity Check

1 steps flagged · score 4.0 of 10

The regime classification and uncertainty bounds depend on σ_stoch, which is fit from the same residuals used to validate it; the 60% uncertainty reduction follows algebraically from the Bayesian update.

  1. fitted input called prediction [Section II D (Eq. 11), Section II E (Eq. 16), Section III E (Fig. 5)]
    "The residual ∆µ is the difference between the predicted state mean and the measurement mean ∆µ = µmeas − µpred (10) and the variance of the residual combines the variances of the prediction and the measurement σ2∆µ = σ2pred + σ2meas ≈ σ2posterior + σ2stoch + σ2meas (11) ... Therefore, we use this analytic equation as a starting point to fit the intrinsic stochasticity by maximizing the log-likelihood numerically via Stochastic Gradient Descent. ... The histograms show that the residuals match the distribution (green curve) that one would expect from Eq. (11)."

    The σ2_posterior inside Eq. (11) is itself determined by σ2_stoch through the Bayesian update, Eq. (6) (with stationary limit Eq. (8)); the residuals in Eq. (10) come from µ_pred, whose parameters α, β are fit on the same N=20 window by minimizing those same residuals (Eq. (16)). Thus σ2_stoch is chosen so that the model's own residuals match the model's own variance formula, and the histogram 'validation' against Eq. (11) is a self-consistency check, not an independent test. The subsequent regime classification (σ2_meas vs 2σ2_stoch) and the posterior uncertainty bounds are therefore in-sample consequences of this fitted parameter.

full rationale

The analytical threshold σ2_meas = 2σ2_stoch is a genuine derivation from the Gaussian update equations and is not circular. The measurement noise Σ2_meas is independently calibrated from device parameters in Section III D. The main circularity is that σ2_stoch, which sets the regime and the posterior width, is estimated from residuals of the same local-linear model whose parameters are fit to those residuals, and then the residuals are shown to match the very variance decomposition used in the likelihood. The reported 54-60% uncertainty reduction also follows algebraically from Eq. (6): for any finite prior variance the posterior variance is necessarily smaller than the measurement variance, so the improvement is a mathematical consequence of the update rule rather than an independent empirical observation. The self-citations to Ref. [13] provide the forward model and are not themselves load-bearing in deriving the threshold. Overall the framework has independent mathematical content, but the experimental regime classification and uncertainty bounds are partly in-sample, giving partial circularity rather than full collapse of the derivation.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The framework rests on standard Gaussian/Markov assumptions and a deliberately simple local-linear model; the key parameters sigma_stoch, alpha, beta are fit to the same data they later classify, which is a limitation.

free parameters (5)
  • alpha (Holt level smoothing) = 0.1596 (average)
    Learned by maximizing log-likelihood over sliding window (Eq. 16); controls how fast the estimated state tracks new measurements.
  • beta (Holt trend smoothing) = 0.0080 (average)
    Learned similarly; controls trend adaptation with half-life ~14.4 min.
  • sigma_stoch (intrinsic stochasticity) = Not reported as single value; varies per mode
    Estimated from residuals by maximum likelihood / stochastic gradient descent (Section II D); central to regime classification.
  • sigma_meas (measurement noise) = 0.2 px spot position + 0.6 px model fit
    Calibrated from spot fitting accuracy and a chosen model-fit factor (Section III D); treated as known in the Bayesian update.
  • sliding window N = 20
    Chosen to cover ~3 minutes; affects stationarity and noise estimates.
assumptions (5)
  • domain assumption Independent Gaussian noise for shot-to-shot fluctuations and measurement noise
    Motivated by Central Limit Theorem (Section II), but not empirically verified for all modes.
  • standard math Markov property for the prior state
    Assumed in Section II A; standard for recursive Bayesian filtering.
  • domain assumption Local linear approximation of the deterministic trend
    Eq. (13) valid only when |v_k dt| << sigma_stoch; may fail during strong drifts.
  • ad hoc to paper Diagonal covariance of intrinsic stochasticity (no mode correlations)
    Stated in Section II D for simplicity; yields upper-bound uncertainties.
  • domain assumption Local stationarity within the sliding window
    Needed for fitting alpha, beta, and sigma_stoch over N=20 samples.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Bayesian perspective on single-shot laser characterization." pith.science (2026). https://pith.science/paper/6BP2EDHE

@misc{pith2026250203100,
  author       = {Pith},
  title        = {Pith review of: A Bayesian perspective on single-shot laser characterization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6BP2EDHE}},
  note         = {Machine review of arXiv:2502.03100}
}
read the original abstract

We introduce a Bayesian framework for measuring spatio-temporal couplings (STCs) in ultra-intense lasers that reconceptualizes what constitutes a 'single-shot' measurement. Moving beyond traditional distinctions between single- and multi-shot devices, our approach provides rigorous criteria for determining when measurements can truly resolve individual laser shots rather than statistical averages. This framework shows that single-shot capability is not an intrinsic device property but emerges from the relationship between measurement precision and inherent parameter variability. Implementing this approach with a new measurement device at the ATLAS-3000 petawatt laser, we provide the first quantitative uncertainty bounds on pulse front tilt and curvature. Notably, we observe that our Bayesian method reduces uncertainty by up to 60% compared to traditional approaches. Through this analysis, we reveal how the interplay between measurement precision and intrinsic system variability defines achievable resolution -- insights that have direct implications for applications where precise control of laser-matter interaction is critical.

Figures

Figures reproduced from arXiv: 2502.03100 by the authors.

Figure 1
Figure 1. FIG. 1: Decomposition of measurements into deterministic [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2: Evolution of the normalized posterior standard de [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. At our regime threshold, we find that the Nyquist frequency ωs/2 is attenuated to about 67%. We can roughly quantify the ability to resolve frequencies using the cutoff frequency ωc = arccos((γ 2+2γ−2)/(2γ−2))ωs for which the spectral power decays to ∥H(ωc/ωs)∥ 2 = 1/2. Shot-to-shot dynamics at frequencies above this cut￾off tend to not be resolved and will appear stochastic. The concept of intrinsic stochasticity, … view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: FIG. 4: Single-shot FALCON, consisting of a mosaic band [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5: The evolution of the pulse front tilt (left) and linear-frequency dependent-coma (right) prediction values together with [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6: Update weight and intrinsic stochasticity for PFT [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: FIG. 7: Logarithmic spectra for the prediction, posterior and [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]
Figure 8
Figure 8. Figure 8: FIG. 8: The evolution of the tilt (left) and defocus (right) prediction values together with the intrinsic stochasticity. The zoomed [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]
Figure 9
Figure 9. Figure 9: FIG. 9: The evolution of the linear Astigmatism (top left), Pulse Front Curvature (PFT)/linear defocus (top right), linear [PITH_FULL_IMAGE:figures/full_fig_p013_9.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Information-Optimal Sensing and Control in High-Intensity Laser Experiments

    physics.optics 2025-06 conditional novelty 4.0 of 10

    The authors propose that optimal measurement and control in high-intensity laser experiments should maximize mutual information about task-relevant parameters, unifying their prior diagnostics and optimization work.

Reference graph

Works this paper leans on

31 extracted references · 29 canonical work pages · cited by 1 Pith paper

  1. [1]

    Danson, D

    C. Danson, D. Hillier, N. Hopps, and D. Neely, High Power Laser Science and Engineering 3, e3 (2015)

    work page 2015

  2. [2]

    J. W. Yoon, Y. G. Kim, I. W. Choi, J. H. Sung, H. W. Lee, S. K. Lee, and C. H. Nam, Optica 8, 630 (2021)

    2021

  3. [3]

    Bowlan, P

    P. Bowlan, P. Gabolde, and R. Trebino, Optics Express 15, 10219 (2007)

    work page 2007

  4. [4]

    Bowlan, P

    P. Bowlan, P. Gabolde, M. A. Coughlan, R. Trebino, and R. J. Levis, Journal of the Optical Society of America B 25, A81 (2008)

    work page 2008

  5. [5]

    Gallet, S

    V. Gallet, S. Kahaly, O. Gobert, and F. Qu´ er´ e, Optics Letters 39, 4687 (2014)

    work page 2014

  6. [6]

    Alonso, ´I

    B. Alonso, ´I. J. Sola, ´O. Varela, J. Hern´ andez-Toro, C. M´ endez, J. S. Rom´ an, A. Za ¨ ır, and L. Roso, Jour- nal of the Optical Society of America B 27, 933 (2010)

    work page 2010

  7. [7]

    Alonso, M

    B. Alonso, M. Miranda, ´I. J. Sola, and H. Crespo, Optics Express 20, 17880 (2012)

    work page 2012

  8. [8]

    Alonso, R

    B. Alonso, R. Borrego-Varillas, O. Mendoza-Yero, ´I. J. Sola, J. S. Rom´ an, G. M ´ ınguez-Vega, and L. Roso, Jour- nal of the Optical Society of America B 29, 1993 (2012)

    work page 2012

Show all 31 references
  1. [9]

    Alonso, M

    B. Alonso, M. Miranda, F. Silva, V. Pervak, J. Rauschen- berger, J. S. Rom´ an,´I. J. Sola, and H. Crespo, Applied Physics B 112, 105 (2013)

    2013

  2. [10]

    Miranda, M

    M. Miranda, M. Kotur, P. Rudawski, C. Guo, A. Harth, A. L’Huillier, and C. L. Arnold, Optics Letters 39, 5142 (2014)

    2014

  3. [11]

    Pariente, V

    G. Pariente, V. Gallet, A. Borot, O. Gobert, and F. Qu´ er´ e, Nature Photonics10, 547 (2016)

    2016

  4. [12]

    Borot and F

    A. Borot and F. Qu´ er´ e, Optics express26, 26444 (2018)

    2018

  5. [13]

    Weisse, J

    N. Weisse, J. Esslinger, S. Howard, F. M. Foerster, F. Haberstroh, L. Doyle, P. Norreys, J. Schreiber, S. Karsch, and A. D¨ opp, Optics Express 31, 19733 (2023), ISSN 1094-4087

    2023

  6. [14]

    Smartsev, A

    S. Smartsev, A. Liberman, I. Andriyash, A. Cavagna, A. Flacco, C. Giaccaglia, J. Kaur, J. Monzac, S. Tata, A. Vernier, et al., Optics Letters 49, 1900 (2024)

    2024

  7. [15]

    Gabolde and R

    P. Gabolde and R. Trebino, Optics Express 14, 11460 (2006)

    2006

  8. [16]

    Dorrer and S.-W

    C. Dorrer and S.-W. Bahk, Optics Express 26, 33387 (2018)

    2018

  9. [17]

    Y. G. Kim, J. I. Kim, J. W. Yoon, J. H. Sung, S. K. Lee, and C. H. Nam, Optics Express 29, 19506 (2021)

    2021

  10. [18]

    H. Tang, T. Men, X. Liu, Y. Hu, J. Su, Y. Zuo, P. Li, J. Liang, M. C. Downer, and Z. Li, Light: Science & Applications 11, 244 (2022), ISSN 2047-7538

    2022

  11. [19]

    Howard, J

    S. Howard, J. Esslinger, R. H. W. Wang, P. Norreys, and A. D¨ opp, High Power Laser Science and Engineering11, e32 (2023)

    2023

  12. [20]

    Alonso, A

    B. Alonso, A. D¨ opp, and S. W. Jolly, APL Photonics 9 (2024)

    2024

  13. [21]

    D¨ opp, C

    A. D¨ opp, C. Eberle, S. Howard, F. Irshad, J. Lin, and M. Streeter, High Power Laser Science and Engineering 11, e55 (2023)

    2023

  14. [22]

    Debus, R

    A. Debus, R. Pausch, A. Huebl, K. Steiniger, R. Widera, T. E. Cowan, U. Schramm, and M. Bussmann, Physical Review X 9, 031044 (2019). 11

    2019

  15. [23]

    J. P. Palastro, J. L. Shaw, P. Franke, D. Ramsey, T. T. Simpson, and D. H. Froula, Physical Review Letters124, 134802 (2020)

    2020

  16. [24]

    Caizergues, S

    C. Caizergues, S. Smartsev, V. Malka, and C. Thaury, Nature Photonics 14, 475 (2020)

    2020

  17. [25]

    Vincenti and F

    H. Vincenti and F. Qu´ er´ e, Physical Review Letters108, 113904 (2012)

    2012

  18. [26]

    Qu´ er´ e, H

    F. Qu´ er´ e, H. Vincenti, A. Borot, S. Monchoc´ e, T. J. Ham- mond, K. T. Kim, J. A. Wheeler, C. Zhang, T. Ruchon, T. Auguste, et al., Journal of Physics B: Atomic, Molec- ular and Optical Physics 47, 124004 (2014)

    2014

  19. [27]

    Mohamed and K

    A. Mohamed and K. Schwarz, Journal of geodesy 73, 193 (1999)

    1999

  20. [28]

    C. C. Holt, International journal of forecasting 20, 5 (2004)

    2004

  21. [29]

    Haffa, J

    D. Haffa, J. Bin, M. Speicher, K. Allinger, J. Hart- mann, C. Kreuzer, E. Ridente, T. M. Ostermayr, and J. Schreiber, Scientific Reports 9, 7697 (2019)

    2019

  22. [30]

    Pretzler, A

    G. Pretzler, A. Kasper, and K. Witte, Applied Physics B 70, 1 (2000)

    2000

  23. [31]

    Irshad, C

    F. Irshad, C. Eberle, F. M. Foerster, K. v. Grafenstein, F. Haberstroh, E. Travac, N. Weisse, S. Karsch, and A. D¨ opp, Phys. Rev. Lett. 133, 085001 (2024), URL https://link.aps.org/doi/10. 1103/PhysRevLett.133.085001. 12 Supplemental Material: Additional Modal Plots FIG. 8: T...

    2024

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.