Pith. sign in

REVIEW 4 major objections 7 minor 43 references

Towards real-time surrogate-free Bayesian inversion for neutron reflectometry

T0 review · 4 major / 7 minor · reviewed 2026-08-04 · deepseek-v4-flash

Pith's one-line read By threading automatic differentiation through the exact reflectivity model, this paper makes gradient-based Bayesian inference — Hamiltonian Monte-Carlo and variational inference — practical for neutron reflectometry, cutting uncertainty q

desk verdict Worth reading: exact AD gradients through the Abeles model make HMC and VI practical for NR, and the paper is honest about its limits; but the headline speed gains are not yet secured by the evidence. read the letter →

arxiv 2509.06924 v3 pith:V3JLAGRP submitted 2025-09-08 cs.LG cond-mat.mtrl-sci

classification cs.LGcond-mat.mtrl-sci
keywords neutronreflectometryBayesianinferenceHamiltonianMonteCarlovariationalautomaticdifferentiationinverseproblemsuncertaintyquantificationthin-filmanalysis
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

The paper aims to make Bayesian uncertainty quantification for neutron reflectometry practical by computing exact gradients of the forward specular reflectivity model — the Abeles matrix formalism that maps layer thicknesses, scattering length densities, and roughnesses to the measured reflectivity — with respect to every unknown parameter. This is claimed to be the first time exact gradients through the reflectivity are available, unlocking gradient-based inference schemes that gradient-free MCMC cannot use. On a benchmark quartz film, the posterior-mean fit reaches a chi-squared of 1.22 versus a published 1.32, and the Hamiltonian Monte-Carlo sampler (NUTS) builds effective sample size steadily where a gradient-free adaptive MCMC scheme stalls. On a 59-parameter co-fit of four organic LED devices, variational inference returns an approximate posterior in under 20 seconds per run, putting UQ on the timescale of fast kinetic experiments. The authors argue this makes surrogate machine-learning models unnecessary for fast NR inversion and release an open-source library so the community can build on the approach.

What carries the argument

The load-bearing object is the exact gradient of the specular reflectivity R̂(Q, θ) obtained by automatic differentiation through the Abeles matrix formalism, the closed-form multilayer model that maps thicknesses, scattering length densities, and interfacial roughnesses to reflected intensity. That gradient flows into any error function or into the evidence lower bound (ELBO) of a variational surrogate q(θ; φ), turning NR inversion into the sort of optimisation problem that ADAM, the NUTS Hamiltonian sampler, and reparameterised variational inference already solve efficiently. The single-sample reparameterisation estimate of the ELBO gradient is what makes VI fast, and symplectic leapfrog i

What would settle it

Fit a deliberately misspecified slab model, such as too few layers or a fixed smearing width, to a simulated dataset with known ground truth: if NUTS or VI then report narrow, confident posteriors incompatible with the true SLD profile, the Gaussian-likelihood/slab-model assumption — not the gradient machinery — limits the method. A cheaper check is to re-analyse the quartz data with a likelihood that explicitly models correlations from the known pointwise smearing and see whether the reported posterior widths change materially.

Watch

Extended reading notes

Core claim

For the first time in neutron reflectometry, exact gradients of the specular reflectivity with respect to all slab and instrument parameters are computed, by differentiating through the Abeles matrix formalism and its smearing kernels. Because the gradients are exact and cheap, the full gradient-based Bayesian toolbox applies directly to the physical forward model: Hamiltonian Monte-Carlo (specifically the NUTS variant) converges to well-mixed posteriors within about 2000 samples per chain where a gradient-free sample-adaptive MCMC scheme stalls, and variational inference fits a 59-parameter joint posterior over four organic LED devices in under 20 seconds per run, with all predictive fits b

Load-bearing premise

Everything rests on the assumption that the chosen slab model, with a Gaussian likelihood whose variances are taken directly from the neutron count data, is the true generator of the measured reflectivity; if the reduced data carry correlated errors or the layer model is misspecified, the posterior widths and speed comparisons inherit that error.

Editorial extensions

If this is right

  • Gradient-based optimisation with ADAM on exact gradients fits the quartz benchmark to chi-squared 1.302, slightly below the published best of 1.32, with all layer and instrument parameters free.
  • NUTS reaches well-converged effective sample sizes within roughly 2000 samples per chain where a gradient-free sample-adaptive MCMC scheme loses effective sample size, making HMC a practical replacement for MCMC when accurate UQ is required.
  • Variational inference produces a full approximate posterior over 59 joint parameters, co-fitting four OLED devices with shared instrument parameters, in under 20 seconds per run, bringing UQ to the timescale of fast kinetic NR experiments.
  • VI's mode-seeking tendency to understate posterior variance is acknowledged and demonstrated on a lipid bilayer benchmark against HMC, giving practitioners a concrete picture of the speed-versus-fidelity trade-off.
  • The library makes the kernels, smearing options, gradients, just-in-time compilation, and GPU parallelism available so the demonstrated optimisation and inference schemes can be reproduced and extended.

Reading between the lines

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

  • If the gradient route holds up, it plausibly transfers to any indirect scattering technique with a differentiable forward model — small-angle scattering, ellipsometry, grazing-incidence scattering — since nothing in the method is specific to reflectometry beyond the kernel; a direct test is re-running the same VI/HMC pipeline on those forward models.
  • The seconds-scale runtime opens the door to closed-loop process control and active experimental design at the beamline without surrogates, for example using the VI posterior of one measurement to choose settings for the next; the paper gestures at this via Bayesian optimisation but does not demonstrate it.
  • Because the surrogate q is chosen to be Gaussian, VI will understate uncertainty whenever the true posterior is multimodal or skewed, which the phase-loss degeneracy of reflectometry makes likely; a testable extension is a more flexible surrogate family such as mixtures or normalising flows to recover tails while keeping the runtime short.
  • The posterior differences between OLED devices, such as the thicker oxide inferred for the 140°C-annealed sample, come from a single measurement each; replicate devices would reveal whether those differences are genuine sample variability or slack in the slab-model degrees of freedom.
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 / 7 minor

Summary. The paper presents refjax, an open-source JAX library that computes specular neutron reflectivity from an arbitrary slab model and, via automatic differentiation, exact gradients with respect to model parameters. These gradients are used to run gradient-based optimisation (Adam), Hamiltonian Monte Carlo (NUTS), and variational inference on NR inverse problems. The method is demonstrated on a crystalline quartz film on silicon (D17 instrument) and on a joint fit of four OLED devices (59 parameters), with an additional lipid bilayer benchmark in the Appendix. The authors report a posterior-mean χ² of 1.22 for quartz (better than the published 1.32), an OLED co-fit via VI in under 20 seconds on a 64-core workstation, and an ESS-per-iteration comparison indicating that NUTS converges where a sample-adaptive MCMC sampler appears to stall. The central claim is that exact gradients through the physical forward model provide a step-change in inference speed and sample efficiency compared with gradient-free MCMC, and that VI can deliver approximate UQ on the order of seconds.

Significance. If the central claim holds, the work is a significant practical contribution to NR analysis: it removes the need for surrogate models, provides a public code base (refjax), and opens the door to gradient-based Bayesian methods for a community that still largely uses Levenberg-Marquardt or gradient-free samplers. The paper is generally honest: it discloses the non-physical quartz oxide SLD, notes the mode-seeking variance underprediction of VI, and includes an external benchmark against refnx. The use of exact gradients through the Abeles model is conceptually clean and the released code makes the approach reproducible. However, the quantitative evidence for the headline speed/efficiency gains is currently incomplete. The ESS comparison is not cost-adjusted, and the VI timing has no same-hardware baseline, so the 'order of magnitude' and 'seconds rather than hours' claims are not yet established. The paper would be strengthened by a careful cost-adjusted benchmark on multiple datasets and posterior geometries.

major comments (4)
  1. [Section 3.1, Figure 6] The sample-efficiency comparison is reported as ESS per iteration, not ESS per wall-clock time. The text acknowledges that NUTS's per-sample cost is 'several times higher' than SA-MCMC, but no ratio is given. Because the Abstract claims 'order of magnitude gains in sample-efficiency' and the Introduction claims a 'step-change in inference speed', the comparison must be normalised by computational cost. Please report ESS per second (or per cost-normalised gradient evaluation) on the same hardware, and state the per-sample cost ratio. If the ratio is, say, 5–10×, the ESS-per-iteration advantage shown in Fig. 6 may disappear.
  2. [Section 3.2, OLED VI] The timing claim ('<20 seconds', 'seconds rather than hours') is based on a single run of VI on a 64-core Threadripper with no same-hardware comparison to HMC/NUTS or to a gradient-free sampler on the same 59-parameter problem. The 'hours' appears to come from typical literature experience rather than from measurement. Please provide a side-by-side wall-clock comparison (including total time for all 100 restarts) and report the elapsed time for each method on the same hardware and data. In addition, selecting the best of 100 restarts by maximum noisy ELBO can bias the chosen surrogate posterior; please report the spread of ELBO values across restarts and/or use a common random seed or a deterministic ELBO evaluation for the selection.
  3. [Section 1.2, Eq. (25), Table 1] The reported posterior widths and the VI-versus-HMC comparison are conditional on a Gaussian likelihood with variances taken from the count data and on the assumed slab model (layer count, Gaussian roughness, smearing kernel). The paper does not validate this likelihood against the data. For instance, the quartz posterior mean assigns a native-oxide SLD of 0.323 Å^-2, far from the physical value (~3.4), which the authors attribute to slab-model slack. This indicates sensitivity to the model class. Please add a posterior predictive check (e.g., simulating from the posterior and comparing to the observed data) and, ideally, a simulation-based calibration study to assess whether the reported posterior widths are trustworthy.
  4. [Section 3.1, Figure 6 and convergence diagnostics] The ESS plot is described as 'the effective sample size of a single parameter in a single chain' and no R-hat diagnostics are reported for the full parameter vector. The sample-efficiency claim should be supported by summary ESS and R-hat across all parameters and all chains, not just one trace. Please specify which parameter(s) are plotted, report the convergence diagnostics for the entire model, and ensure the plotted chain/parameter is representative rather than cherry-picked.
minor comments (7)
  1. [Introduction] 'complex multiplayer structures' should be 'multilayer structures'.
  2. [Section 3.1] 'numypro' is a typo for 'numpyro'.
  3. [Table 4] The background initial value '5.0' is inconsistent with its prior U(e^-20, 1×10^-7); presumably the intended value is 5.0×10^-7. Please correct.
  4. [Appendix B] 'D20' and 'H20' should be 'D2O' and 'H2O'.
  5. [Section 1.4] 'variation inference' should be 'variational inference'.
  6. [Figure 6 caption] The caption is ambiguous: it says 'each line is the effective sample size of a single parameter in a single chain'—please clarify which parameters/chains are shown and how many lines appear.
  7. [Abstract] The claim 'for the first time in NR, exact gradients through the reflectivity are computed' is stated as a fact. The authors may wish to soften this to 'to the best of our knowledge' and ensure the literature search covers adjacent fields (X-ray reflectometry, optical thin films) where AD has been used.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the forward model is the standard Abeles formalism, inference uses standard numpyro samplers, and benchmarks are external datasets; speed claims are benchmarking concerns, not circular reductions.

full rationale

The derivation chain is self-contained. The forward reflectivity model is the standard Abeles matrix formalism ('The Abeles matrix formalism [23] provides a physical model of the reflectivity'), not a parameter fitted to the data, and the claimed novelty is exact automatic-differentiation gradients through this fixed model. The inference algorithms (NUTS, Sample Adaptive MCMC, ADAM, VI) are standard implementations from numpyro and JAX, not bespoke constructions that encode the conclusions. The empirical comparisons use external datasets: D17 quartz data ('The reflectivity data were collected on the D17 neutron reflectivity instrument at the ILL'), ISIS OLED data, and a refnx lipid benchmark ('this system makes an ideal benchmark for comparison of new fitting methodologies'). Warm-starting HMC at the Adam optimum is explicitly disclosed ('The chain was initialised prior to the burn-in at the optimal value found by gradient descent'), and the reported χ²=1.22 is a same-data fit-quality measure, not an out-of-sample prediction; this is a benchmarking weakness but not circularity. The admitted limitations — higher per-sample NUTS cost and VI variance underprediction (Appendix C) — are transparent trade-offs, not hidden reductions of outputs to inputs. No load-bearing self-citation or imported uniqueness/ansatz is present; even the reference to [38] is only a data source and published baseline, not an argumentative premise.

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

The demonstrations rest on standard mathematical machinery plus domain assumptions: the Abeles forward model and slab parametrizations, the Gaussian count-data likelihood, hand-set priors and algorithm hyperparameters, and the Gaussian surrogate for VI. The slab-model parameters are the outputs of inference rather than hidden inputs. No new physical entities are introduced; refjax is software, not an entity.

free parameters (5)
  • VI surrogate initialization and schedule = sigma_0=0.1, learning rate 0.05, 2000 iterations, 100 restarts
    Hand-chosen hyperparameters (Section 3.2). The best-of-100 ELBO selection can bias the reported approximate posterior, and the 20-second timing is per run on one workstation.
  • HMC/NUTS chain budget = 1000 burn-in samples, 10 chains of 2000 samples
    Chosen by the authors (Section 3.1). The ESS comparison and the 'fraction of samples required' claim depend on these totals.
  • Slab model layer counts and prior bounds = quartz: 2 layers; OLED: 4 layers; uniform priors in Tables 1-4
    Modeling choices made by the analyst. The priors are described as weak knowledge but are hand-set ranges and directly shape posterior widths.
  • OLED shared smearing parameter = posterior 5.56±0.06 percent
    Fitted as an unknown shared instrument parameter (Table 6). The functional form, constant Gaussian smearing, is assumed, not validated against measured resolution.
  • LED per-device scale and shared background = scale: 0.954, 1.02, 0.998, 0.963; background: 4.99e-07
    Legitimate fitted instrument parameters (Table 6), listed because they are fitted to data and shape the likelihood; they do not affect the central speed claim directly.
assumptions (5)
  • domain assumption The Abeles matrix formalism exactly describes specular reflectivity for the slab models used
    Adopted as the forward model throughout (Section 1.2). Standard in NR, but layer count and roughness parametrization are fixed by the analyst.
  • domain assumption The likelihood p(R|theta) is Gaussian with variances taken from measured count data
    Stated explicitly in Section 1.2 as an approximation. All reported posterior widths and chi-squared values inherit this choice; correlated errors from data reduction are ignored.
  • domain assumption A Gaussian surrogate on an unconstrained space adequately represents the posterior for VI purposes
    Used in Section 3.2 and Appendix B. The authors discuss the resulting mode-seeking variance underestimation in Appendix C.
  • standard math NUTS default adaptation (step size and mass matrix) is a correct and robust HMC implementation
    Imported from numpyro (ref 37); the paper relies on its defaults for the ESS comparison in Section 3.1.
  • domain assumption The prior bounds in Tables 1-4 contain all physically plausible parameter values
    Uniform priors with hand-set ranges; the resulting posteriors are conditional on these ranges, including the roughness bounds used around the oxide layer.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards real-time surrogate-free Bayesian inversion for neutron reflectometry." pith.science (2026). https://pith.science/paper/V3JLAGRP

@misc{pith2026250906924,
  author       = {Pith},
  title        = {Pith review of: Towards real-time surrogate-free Bayesian inversion for neutron reflectometry},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/V3JLAGRP}},
  note         = {Machine review of arXiv:2509.06924}
}
read the original abstract

Neutron reflectometry (NR) is a key enabling technology for many areas of scientific development. Although the forward reflectivity model is well-known, inferring the physical properties of a sample from NR data requires the solution of an inverse problem. Increasingly, beamline scientists are using NR in fast kinetic configurations and probing highly-complex structures and interfaces, introducing significant uncertainty. Existing uncertainty quantification (UQ) approaches in NR, such as Markov-Chain Monte-Carlo (MCMC), suffer from poor sample efficiency and slow convergence times. Recently, surrogate machine learning models have been proposed as an alternative. However, physical intuition is lost when replacing governing equations with fast surrogates. Instead, we propose a rapid, surrogate-free Bayesian inversion for NR. Our approach offers a step-change in inference speed and efficiency. For the first time in NR, exact gradients through the reflectivity are computed, enabling highly performant gradient-based inference schemes: Hamiltonian Monte-Carlo offers significant advances in sample efficiency compared to MCMC. Variational inference enables approximate UQ on the order of seconds rather than hours. We demonstrate state-of-the-art performance on a thick oxide quartz film, and robust co-fitting performance in the high complexity regime of organic LED multilayer devices. Additionally, we provide an open-source library of reflectometry kernels in the python language.

Figures

Figures reproduced from arXiv: 2509.06924 by the authors.

Figure 1
Figure 1. The inverse problem at the heart of NR. Both the sample and the reflectometer instrument are described by [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Automatic-differentiation back propagates the gradient of the error function through the forward reflectivity [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Optimisation of quartz thin film NR dataset fit using gradient decent with the ADAM optimiser [ [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figures from the paper (11 more)
Figure 4
Figure 4. Figure 4: Inference of quartz NR model by Hamiltonian Monte-Carlo. All parameters remain free to fit. It is notable [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: Posterior SLD profile and posterior marginal parameter distributions for the NR parameters of the quartz film, [PITH_FULL_IMAGE:figures/full_fig_p012_5.png]
Figure 6
Figure 6. Figure 6: Comparison of effective sample sizes per iteration between a Hamiltonian Monte-Carlo (NUTS) scheme [PITH_FULL_IMAGE:figures/full_fig_p013_6.png]
Figure 7
Figure 7. Figure 7: (a) Results of the inference for the LED devices. Note that the reflectivity curves are spaced for visibility on [PITH_FULL_IMAGE:figures/full_fig_p015_7.png]
Figure 8
Figure 8. Figure 8: Posterior distribution identified over the slab model parameters of the crystalline quartz sample. [PITH_FULL_IMAGE:figures/full_fig_p020_8.png]
Figure 9
Figure 9. Figure 9: Posterior distribution identified over the slab model parameters of the LED devices. [PITH_FULL_IMAGE:figures/full_fig_p021_9.png]
Figure 10
Figure 10. Figure 10: Identified VI posterior distributions of NR parameters for the LED data. (a) Background level (shared [PITH_FULL_IMAGE:figures/full_fig_p023_10.png]
Figure 11
Figure 11. Figure 11: Slab model of a bilayer Lipid. Each slab on the right is charaterised by a single SLD and thickness. Each [PITH_FULL_IMAGE:figures/full_fig_p024_11.png]
Figure 12
Figure 12. Figure 12: Results of the inference for the lipid bilayer – Predicted reflectometry curves. (a) Posterior estimated [PITH_FULL_IMAGE:figures/full_fig_p025_12.png]
Figure 13
Figure 13. Figure 13: Results of the inference for the lipid bilayer – SLD profiles. (a) Posterior estimated by a variational inference [PITH_FULL_IMAGE:figures/full_fig_p026_13.png]
Figure 14
Figure 14. Figure 14: Posterior corner plot of identified parameters from both the NUTS (green) and VI (orange) inference [PITH_FULL_IMAGE:figures/full_fig_p027_14.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 38 canonical work pages

  1. [1]

    M. W. Skoda. Recent developments in the application of x-ray and neutron reflectivity to soft-matter systems. Current Opinion in Colloid & Interface Science, 2019

  2. [2]

    B. J. Kirby, P. A. Kienzle, B. B. Maranville, N. F. Berk, J. Krycka, F. Heinrich, and C. Majkrzak. Phase-sensitive specular neutron reflectometry for imaging the nanometer scale composition depth profile of thin-film materials. Current Opinion in Colloid & Interface Science, 2012

  3. [3]

    Junghans, E

    A. Junghans, E. B. Watkins, R. D. Barker, S. Singh, M. J. Waltman, H. L. Smith, L. Pocivavsek, and J. Majewski. Analysis of biosurfaces by neutron reflectometry: From simple to complex interfaces.Biointerphases, 2015. 17 Towards real-time surrogate-free Bayesian inversion for neutron reflectometryPREPRINT

  4. [4]

    C. F. Majkrzak, J. W. Cable, J. Kwo, M. Hong, T. Shih, and J. V . Waszczak. Neutron diffraction studies of fe/cr superlattices.Journal of Applied Physics, 1988

  5. [5]

    Liley, J

    J. Liley, J. Penfold, R. K. Thomas, I. Tucker, J. Petkov, P. Stevenson, I. Banat, R. Marchant, M. Rudden, and J. Webster. The performance of surfactant mixtures at low temperatures.Journal of colloid and interface science, 2019

  6. [6]

    Pabst, N

    G. Pabst, N. Kuˇcerka, M.-P. Nieh, M. Rheinstädter, and J. Katsaras. Applications of neutron and x-ray scattering to the study of biologically relevant model membranes.Chemistry and Physics of Lipids, 2010

  7. [7]

    L. A. Clifton, S. A. Holt, A. V . Hughes, E. L. Daulton, W. Arunmanee, F. Heinrich, S. Khalid, D. Jefferies, T. R. Charlton, J. R. Webster,et al. An accurate in vitro model of the e. coli envelope.Angewandte Chemie International Edition, 2015

  8. [8]

    Penfold and R

    J. Penfold and R. Thomas. The application of the specular reflection of neutrons to the study of surfaces and interfaces.Journal of Physics: Condensed Matter, 1990

Show all 43 references
  1. [9]

    A. R. Nelson and S. W. Prescott. refnx: neutron and x-ray reflectometry analysis in python.Applied Crystallogra- phy, 2019

  2. [10]

    J. Woollam. Completeease, 2017

  3. [11]

    Levenberg

    K. Levenberg. A method for the solution of certain non-linear problems in least squares.Quarterly of Applied Mathematics, 1944

  4. [12]

    D. W. Marquardt. An algorithm for least-squares estimation of nonlinear parameters.Journal of the Society for Industrial and Applied Mathematics, 1963

  5. [13]

    Nocedal and S

    J. Nocedal and S. J. Wright.Numerical Optimization. Springer Series in Operations Research and Financial Engineering. Springer, 2nd edition, 2006

  6. [14]

    Storn and K

    R. Storn and K. Price. Differential evolution - a simple and efficient heuristic for global optimization over continuous spaces.Journal of Global Optimization, 1997

  7. [15]

    Eberhart and J

    R. Eberhart and J. Kennedy. A new optimizer using particle swarm theory. InMHS’95. Proceedings of the sixth international symposium on micro machine and human science. IEEE, 1995

  8. [16]

    Metropolis, A

    N. Metropolis, A. W. Rosenbluth, M. N. Rosenbluth, A. H. Teller, and E. Teller. Equation of state calculations by fast computing machines.The Journal of Chemical Physics, 1953

  9. [17]

    W. K. Hastings. Monte carlo sampling methods using Markov chains and their applications.Biometrika, 1970

  10. [18]

    Skilling

    J. Skilling. Nested sampling.Bayesian inference and maximum entropy methods in science and engineering, 2004

  11. [19]

    Doucet, R

    M. Doucet, R. K. Archibald, and W. T. Heller. Machine learning for neutron reflectometry data analysis of two-layer thin films.Machine Learning: Science and Technology, 2021

  12. [20]

    Hinderhofer, A

    A. Hinderhofer, A. Greco, V . Starostin, V . Munteanu, L. Pithan, A. Gerlach, and F. Schreiber. Machine learning for scattering data: strategies, perspectives and applications to surface scattering.Applied Crystallography, 2023

  13. [21]

    Pithan, V

    L. Pithan, V . Starostin, D. Mare ˇcek, L. Petersdorf, C. Völter, V . Munteanu, M. Jankowski, O. Konovalov, A. Gerlach, A. Hinderhofer,et al. Closing the loop: autonomous experiments enabled by machine-learning-based online data analysis in synchrotron beamline environments.Sy...

  14. [22]

    Starostin, M

    V . Starostin, M. Dax, A. Gerlach, A. Hinderhofer, Á. Tejero-Cantero, and F. Schreiber. Fast and reliable probabilistic reflectometry inversion with prior-amortized neural posterior estimation.Science Advances, 2025

  15. [23]

    Abelès.La théorie générale des couches minces

    F. Abelès.La théorie générale des couches minces. PhD thesis, Université de Paris, 1949

  16. [24]

    D. P. Kingma. Adam: A method for stochastic optimization.arXiv preprint arXiv:1412.6980, 2014

  17. [25]

    G. Hinton. Lecture 6e rmsprop: Divide the gradient by a running average of its recent magnitude. https: //www.cs.toronto.edu/~tijmen/csc321/slides/lecture_slides_lec6.pdf, 2012. Coursera: Neural Networks for Machine Learning. 18 Towards real-time surrogate-free Bayesian invers...

  18. [26]

    Betancourt

    M. Betancourt. A conceptual introduction to Hamiltonian Monte Carlo.arXiv preprint arXiv:1701.02434, 2017

  19. [27]

    D. M. Blei, A. Kucukelbir, and J. D. McAuliffe. Variational inference: A review for statisticians.Journal of the American statistical Association, 2017

  20. [28]

    Hairer, M

    E. Hairer, M. Hochbruck, A. Iserles, and C. Lubich. Geometric numerical integration.Oberwolfach Reports, 2006

  21. [29]

    M. D. Hoffman, A. Gelman,et al. The no-u-turn sampler: adaptively setting path lengths in Hamiltonian Monte Carlo.J. Mach. Learn. Res., 2014

  22. [30]

    M. I. Jordan, Z. Ghahramani, T. S. Jaakkola, and L. K. Saul. An introduction to variational methods for graphical models.Machine Learning, 1999

  23. [31]

    M. D. Hoffman, D. M. Blei, C. Wang, and J. Paisley. Stochastic variational inference.the Journal of machine Learning research, 2013

  24. [32]

    Yin and M

    M. Yin and M. Zhou. Semi-implicit variational inference. InInternational conference on machine learning. PMLR, 2018

  25. [33]

    D. P. Kingma and M. Welling. Auto-encoding variational bayes. InInternational Conference on Learning Representations (ICLR), 2014

  26. [34]

    Bradbury, R

    J. Bradbury, R. Frostig, P. Hawkins, M. J. Johnson, C. Leary, D. Maclaurin, G. Necula, A. Paszke, J. VanderPlas, S. Wanderman-Milne, and Q. Zhang. JAX: composable transformations of Python+NumPy programs, 2018

  27. [35]

    Rader, T

    J. Rader, T. Lyons, and P. Kidger. Optimistix: modular optimisation in jax and equinox.arXiv preprint arXiv:2402.09983, 2024

  28. [36]

    Babuschkin, K

    DeepMind, I. Babuschkin, K. Baumli, A. Bell, S. Bhupatiraju, J. Bruce, P. Buchlovsky, D. Budden, T. Cai, A. Clark, I. Danihelka, A. Dedieu, C. Fantacci, J. Godwin, C. Jones, R. Hemsley, T. Hennigan, M. Hessel, S. Hou, S. Kapturowski, T. Keck, I. Kemaev, M. King, M. Kunesch, L....

  29. [37]

    D. Phan, N. Pradhan, and M. Jankowiak. Composable effects for flexible and accelerated probabilistic program- ming in numpyro.arXiv preprint arXiv:1912.11554, 2019

  30. [38]

    Gutfreund, T

    P. Gutfreund, T. Saerbeck, M. A. Gonzalez, E. Pellegrini, M. Laver, C. Dewhurst, and R. Cubitt. Towards general- ized data reduction on a chopper-based time-of-flight neutron reflectometer.Journal of Applied Crystallography, 2018

  31. [39]

    A. Hajian. Efficient cosmological parameter estimation with hamiltonian monte carlo technique.Physical Review D—Particles, Fields, Gravitation, and Cosmology, 2007

  32. [40]

    M. Tinto. Gravitational wave searches with pulsar timing arrays: Cancellation of clock and ephemeris noises. Physical Review D, 2018

  33. [41]

    M. Zhu. Sample Adaptive MCMC.Advances in Neural Information Processing Systems, 2019

  34. [42]

    Lanfear, X

    R. Lanfear, X. Hua, and D. L. Warren. Estimating the effective sample size of tree topologies from Bayesian phylogenetic analyses.Genome biology and evolution, 2016

  35. [43]

    B. Ru, M. A. Osborne, M. McLeod, and D. Granziol. Fast information-theoretic Bayesian optimisation. In International Conference on Machine Learning. PMLR, 2018. 19 Towards real-time surrogate-free Bayesian inversion for neutron reflectometryPREPRINT A Posterior distributions P...

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.