Pith. sign in

REVIEW 3 major objections 6 minor 54 references

SECRET: Stochasticity Emulator for Cosmic Ray Electrons

T0 review · 3 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read A neural density estimator now stands in for expensive Monte Carlo simulations of cosmic-ray electron spectra, providing exact likelihoods and fast sampling with few-percent accuracy.

desk verdict Useful, honest emulator for stochastic CRE spectra; the 'exact likelihood' claim oversells it and joint fidelity is unvalidated, but the core interpolation and sample-generation claims hold up. read the letter →

arxiv 2501.06011 v2 pith:IWSVM43L submitted 2025-01-10 astro-ph.HE astro-ph.GAphysics.comp-ph

Nikolas Frediani , Michael Krämer , Philipp Mertsch , Kathrin Nippel (RWTH Aachen) This is my paper · ORCID
classification astro-ph.HEastro-ph.GAphysics.comp-ph
keywords cosmic-rayelectronsstochasticsourceensemblesneuraldensityestimationMADEmaskedautoregressivemodelMonteCarlosimulationemulatortransportsimulation-basedinference
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 introduces SECRET, a neural-network emulator that replaces expensive Monte Carlo simulations of cosmic-ray electron spectra with a fast conditional density model. The central claim is that SECRET can evaluate the exact joint likelihood of a measured spectrum and generate new stochastic spectra in seconds, with few-percent accuracy on the distribution's quantiles over most of the covered parameter space. This matters because the stochasticity of nearby cosmic-ray sources makes the spectrum a non-Gaussian random field; previously only samples from the Monte Carlo ensemble were available, and parameter studies required rerunning the simulation for every parameter change. If the claim holds, parameter inference against data becomes tractable and cheap.

What carries the argument

The central object is the Masked Autoencoder for Distribution Estimation (MADE), a neural network whose weight masks enforce an autoregressive factorization of the joint density, $p(\psi_1,\ldots,\psi_{19})=\prod_i p(\psi_i|\psi_{<i})$. Each conditional is a mixture of Gaussians whose means, widths, and weights are network outputs, which lets the model represent the heavy-tailed, non-Gaussian intensity distributions. SECRET extends MADE by inserting the five physical parameters as additional input dimensions in the autoregressive ordering, so that spectra are conditioned on the parameters exactly, and by training on a hypercubic grid of 6,750 parameter combinations with $10^4$ realizations each. The construction converts an expensive sampling problem into a single trained density estimator that can be queried for likelihoods or used to draw samples in seconds.

What would settle it

Pick a parameter point in the interior of the trained grid, run an independent Monte Carlo set with far more realizations than the training set (e.g., $10^5$), and compare SECRET's predicted likelihood, quantiles, and tail occupancy against those simulations. If the deviations exceed the few-percent level, or if the likelihood surface disagrees with a dedicated Monte Carlo run at an off-grid point, the emulation claim fails.

Watch

Extended reading notes

Core claim

SECRET is a masked autoregressive density estimator (MADE) trained on 67.5 million Monte Carlo spectra spanning a five-dimensional grid of supernova rate, source spectral index, cutoff energy, and diffusion-coefficient parameters. By factorizing the 19-energy-bin joint distribution into conditional probabilities and modeling each conditional as a mixture of Gaussians, the network learns the full joint distribution, including the non-linear correlations between energy bins that are the signature of individual nearby sources. The paper demonstrates that for a fixed parameter set the learned distribution matches the simulated marginals and pairwise correlations, and that the conditioned version interpolates across the grid with median quantile deviations below about 0.05 in log-flux over most of the space. Failures are localized to the low edges of the supernova rate and diffusion-coefficient ranges. The authors conclude that SECRET can evaluate exact likelihoods and generate samples about $10^{4}$ times faster than the Monte Carlo code, making it an emulator suitable for simulation-based inference.

Load-bearing premise

The emulator is only as good as the Monte Carlo transport model that produced its training spectra; if that model's assumptions about diffusion, energy losses, and source distribution do not describe the real Galaxy, SECRET will accurately emulate the wrong distribution.

Editorial extensions

If this is right

  • Exact likelihood evaluation for observed cosmic-ray electron spectra becomes practical, enabling simulation-based inference over source and transport parameters without rerunning Monte Carlo simulations.
  • Generating new spectra is accelerated by roughly four orders of magnitude, making large ensembles or finely sampled parameter scans cheap enough for routine use.
  • Because the conditioning is exact, users can interpolate continuously between the grid points in the trained volume, subject to the identified validity boundaries.
  • The emulator provides a direct estimate of the stochastic variance expected at each energy bin under a given physics model, sharpening interpretations of features like the TeV break.
  • The localization of failures to low supernova rates and low diffusion coefficients defines an explicit region of validity that future users can respect.

Reading between the lines

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

  • A similar autoregressive conditioner could be trained for other source populations, such as pulsar wind nebulae, or extended to include position-dependent diffusion inferred from gamma-ray halos; the paper names these as future directions but does not implement them.
  • The quantile-based accuracy metric says little about the far tails of the distribution, where training samples are sparse; a dedicated tail-focused test would be needed before trusting extreme likelihood values or rare high-flux events.
  • Because SECRET returns a continuous density, it could be embedded in hierarchical Bayesian analyses of multi-messenger data, replacing the Monte Carlo forward model inside Markov-chain samplers.
  • Disagreement between an observed spectrum and the learned ensemble at a claimed parameter point could serve as an anomaly detector for physics beyond the assumed transport and source model.
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

3 major / 6 minor

Summary. The paper introduces SECRET, a masked autoregressive density estimator (MADE) with Gaussian-mixture conditionals, trained to emulate the joint probability distribution of stochastic cosmic-ray electron (CRE) spectra generated by a Monte Carlo model of discrete SNR sources. The authors first validate a single-point MADE for fixed transport parameters, then extend it to SECRET, which conditions on five transport parameters (SN rate, cutoff energy, injection index, diffusion index, and diffusion normalization). They train on a hypercubic grid of 6750 parameter combinations with 10^4 realizations each, report few-percent quantile accuracy over most of the parameter space, identify localized failure regions at the lowest values of R_SN, δ, and κ0, and describe an offset-grid interpolation test. The trained model and code are released publicly. The central claims are that SECRET can evaluate likelihoods and generate samples very efficiently, serving as an emulator for the Monte Carlo simulations.

Significance. If the joint density emulator performs as claimed, it is a useful contribution: it makes likelihood-based inference and fast sample generation feasible for stochastic CRE models, where previously only Monte Carlo samples or approximate copula constructions were available. The paper is transparent that SECRET is a fit to Monte Carlo outputs rather than a first-principles prediction, and it explicitly identifies regions of parameter space where the emulator fails. The release of documented code and pretrained weights is a concrete strength, as is the use of a discriminator for architecture selection in the single-point case. The offset-grid interpolation test is a good idea and, if properly documented, would provide independent support for the interpolation claim. However, the validation of SECRET's joint distribution is currently incomplete, which bears directly on the 'exact likelihood' claim.

major comments (3)
  1. [Sec. 3.3, Figs. 9-10; Sec. 4] The validation of SECRET is entirely in terms of per-energy quantile differences ΔQ_q(ψ(E)) for the 1D marginals. The paper's central claim in Sec. 4 that SECRET 'can evaluate exact likelihoods' requires reproducing the 19-dimensional joint distribution, including the correlation structure between energy bins. No SECRET-specific check of pairwise 2D marginals (the pairwise comparison in Fig. 11 is for the single-point MADE only), of the copula/dependence structure, or of the log-likelihood of held-out MC samples is reported. Since the model is a Gaussian-mixture MADE with K=10 and the authors acknowledge imperfect tail learning (Sec. 3.2.2), matching marginals to a few percent does not imply matching the joint distribution. This gap is load-bearing for the 'exact likelihood' and 'emulator' claims; please add a SECRET-specific joint-distribution diagnostic, such as held-out MC log-likelihood compared to a product-of-marginals baseline, or 2D marginal/energy-bin correlation checks at representative parameter points.
  2. [Sec. 3.3, Figs. 9-10] The reported quantile deviations are presented without any accounting for the sampling uncertainty of the reference MC ensembles, which contain N_realisations = 10^4 spectra per grid point. The empirical quantiles Q_q^sim are estimates whose standard error scales with sqrt(q(1-q)/N)/f(Q_q); for the outer quantiles of the heavy-tailed ψ distributions this error can be comparable to the deviations shown. Without error bars on the MC quantiles, or a comparison of SECRET's errors against the MC-vs-MC variability, the 'few-percent accuracy' statement and the identification of failure regions in Figs. 9-10 are not fully calibrated.
  3. [Sec. 3.3, offset-grid interpolation test] The diagonally offset-grid validation is described only in prose ('errors are typically below ΔQ_q(ψ(E)) ≲ 0.05'), with no figure, table, or exact specification of the offset grid (number of points, offset magnitude, which parameter combinations were simulated). Since this test is the primary evidence for the interpolation capability that distinguishes SECRET from a single-point emulator, it should be documented in a reproducible way, e.g., by showing the error distribution on the offset grid and stating the construction explicitly.
minor comments (6)
  1. [Eq. (1.1) and text] The factorisation in Eq. (1.1) conditions each factor on the following (higher-index) variables, while the surrounding text describes conditioning on lower-index variables, p(ϕ_i|ϕ_{i-1},...ϕ_1); please align the notation with Eq. (3.1).
  2. [Sec. 3.2.1 vs Sec. 2.2] The single-point MADE dataset uses an energy grid of 10^{2.4}-10^{4.2} GeV with 19 bins, whereas the SECRET setup uses 10^{1.5}-10^{4.5} GeV; the relation between these two grids is not explained and should be clarified.
  3. [Fig. 8 and Fig. 10 captions] In the parameter tuples, the unit of κ0 is given as 10^{28} GeV; it should be 10^{28} cm^2 s^{-1}.
  4. [Sec. 3.3, Fig. 9/10 captions] The phrase 'The first and last bins are overflow bins' is unclear in the context of the histograms; please define what 'overflow' means here.
  5. [Sec. 4] The wording 'exact likelihoods' is too strong for a finite-capacity neural density estimator trained on a finite MC sample; consider 'approximate likelihoods' or 'fast likelihood evaluation' throughout.
  6. [Sec. 3.2.2] The classifier accuracy of 62% is described as 'a lower bound on the achievable separability', but the statement would benefit from noting that the discriminator is a specific, non-optimal architecture and the metric is relative rather than absolute.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: SECRET is an openly trained emulator of Monte Carlo outputs, and its interpolation claim is tested on a diagonal offset grid independent of the training grid.

full rationale

SECRET is a surrogate density model fitted to MC-generated spectra, and the paper is transparent about this: it trains on 6750 grid points and evaluates quantile errors against simulations, including a diagonally offset grid that was not part of the training set (Sec. 3.3). The central claim is interpolation accuracy of an emulator, not a first-principles prediction, so no fitted input is relabeled as a prediction. The transport solution follows Ref. [20] and the source distribution follows Ref. [42], both co-authored by P. Mertsch, but those citations supply the physical forward model that generates the training data; they do not by themselves establish the emulator's few-percent quantile accuracy, which rests on the paper's held-out comparisons. The phrase 'evaluate exact likelihoods' (Sec. 4) is stronger than what the 1D quantile validation supports—no SECRET-specific check of energy-bin correlations or held-out log-likelihood is reported—but that is a validation gap and correctness risk, not a circular reduction. No equation defines the target distribution in terms of the fitted model, and no uniqueness theorem is imported from the authors' prior work. The derivation chain is therefore self-contained with respect to circularity.

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

The paper contributes an emulator, not a first-principles derivation. The physical content is inherited from the Monte Carlo model of Ref. [20]; the ML model adds architecture choices and trained weights as fitted components. No new physical entities are introduced.

free parameters (4)
  • tmax = 7 Myr
    Maximum source age in simulations, chosen so that completeness down to 10^1.5 GeV is achieved; it directly sets the number of simulated sources Nsrc = RSN * tmax and thus the stochastic realization properties.
  • number of Gaussian mixture components K = 10
    Architecture choice for MADE output, selected by the authors' classifier-based model search; the central accuracy claim depends on this.
  • hidden layer size = 200 nodes
    Best performing single hidden layer size for the single-point MADE; chosen by performance on a discriminator.
  • SECRET training epochs = 22
    Early-stopped by observing the loss begin to rise; affects final network weights and thus emulator accuracy.
assumptions (4)
  • domain assumption Diffusion-only transport equation with isotropic homogeneous scalar diffusion and continuous energy losses
    Used in Section 2.1 to derive the Green's function; neglects convection, reacceleration, secondaries, and solar modulation.
  • domain assumption Source population is supernova remnants with identical power-law injection spectra with exponential cutoff
    Equation (2.3); all sources share gamma and Ecut, which is a simplification of real source diversity.
  • domain assumption Stochastic sampling: source ages uniform in [0, tmax], distances drawn from Ahlers et al. spiral-arm distribution, causal cut applied
    Section 2.1, dataset construction; the resulting ensemble defines the target distribution that SECRET learns.
  • standard math Autoregressive factorization and Gaussian mixture conditional densities are sufficient to represent the true joint distribution
    Section 3.1; the MADE architecture assumes the chain-rule factorization and that each conditional is a mixture of K Gaussians.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SECRET: Stochasticity Emulator for Cosmic Ray Electrons." pith.science (2026). https://pith.science/paper/IWSVM43L

@misc{pith2026250106011,
  author       = {Pith},
  title        = {Pith review of: SECRET: Stochasticity Emulator for Cosmic Ray Electrons},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IWSVM43L}},
  note         = {Machine review of arXiv:2501.06011}
}
read the original abstract

The spectrum of cosmic-ray electrons depends sensitively on the history and spatial distribution of nearby sources. Given our limited observational handle on cosmic-ray sources, any model remains necessarily probabilistic. Previously, predictions were performed in a Monte Carlo fashion, summing the contributions from individual, simulated sources to generate samples from the statistical ensemble of possible electron spectra. Such simulations need to be re-run if the cosmic-ray transport parameters (e.g. diffusion coefficient, maximum energy) are changed, rendering any parameter study computationally expensive. In addition, a proper statistical analysis of observations and comparison with such probabilistic models requires the joint probability distribution of the full spectrum instead of only samples. Note that parametrising this joint distribution is rendered difficult by the non-Gaussian statistics of the cosmic-ray fluxes. Here, we employ machine learning to compute the joint probability distribution of cosmic-ray electron fluxes. Specifically, we employ masked autoregressive density estimation (MADE) for a representation of the high-dimensional joint probability distribution. In a first step, we train the network on a Monte Carlo simulation for a fixed set of transport parameters, thus significantly accelerating the generation of samples. In a second step, we extend this setup to SECRET (Stochasticity Emulator for Cosmic Ray Electrons), allowing to reliably interpolate over the space of transport parameters. We make the MADE and SECRET codes available at https://git.rwth-aachen.de/pmertsch/secret .

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

54 extracted references · 23 canonical work pages

  1. [1]

    Gabici, C

    S. Gabici, C. Evoli, D. Gaggero, P. Lipari, P. Mertsch, E. Orlando et al.,The origin of Galactic cosmic rays: challenges to the standard paradigm, Int. J. Mod. Phys. D28 (2019) 1930022 [1903.11584]

    arXiv 2019

  2. [2]

    Particle Data Group collaboration, Review of particle physics, Phys. Rev. D110 (2024) 030001

    2024

  3. [3]

    Ginzburg and S.I

    V.L. Ginzburg and S.I. Syrovatskii,The Origin of Cosmic Rays(1964)

    work page 1964

  4. [4]

    Potgieter,Solar Modulation of Cosmic Rays, Living Rev

    M. Potgieter,Solar Modulation of Cosmic Rays, Living Rev. Solar Phys.10 (2013) 3 [1306.4421]

    arXiv 2013

  5. [5]

    Breaks in interstellar spectra of positrons and electrons derived from time-dependent AMS data

    A. Vittino, P. Mertsch, H. Gast and S. Schael,Breaks in interstellar spectra of positrons and electrons derived from time-dependent AMS data, Phys. Rev. D100 (2019) 043007 [1904.05899]

    work page Pith review arXiv 2019

  6. [6]

    Kuhlen and P

    M. Kuhlen and P. Mertsch,Time-dependent AMS-02 electron-positron fluxes in an extended force-field model, Phys. Rev. Lett.123 (2019) 251104 [1909.01154]

    arXiv 2019

  7. [7]

    DAMPE collaboration, Direct detection of a break in the teraelectronvolt cosmic-ray spectrum of electrons and positrons, Nature 552 (2017) 63 [1711.10981]

    arXiv 2017

  8. [8]

    Fermi-LAT collaboration, Cosmic-ray electron-positron spectrum from 7 GeV to 2 TeV with the Fermi Large Area Telescope, Phys. Rev. D95 (2017) 082007 [1704.07195]

    arXiv 2017

Show all 54 references
  1. [9]

    AMS collaboration, The Alpha Magnetic Spectrometer (AMS) on the international space station: Part II — Results from the first seven years, Phys. Rept. 894 (2021) 1

    2021

  2. [10]

    O. Adriani et al.,Extended Measurement of the Cosmic-Ray Electron and Positron Spectrum from 11 GeV to 4.8 TeV with the Calorimetric Electron Telescope on the International Space Station, Phys. Rev. Lett.120 (2018) 261102 [1806.09728]

    2018 arXiv

  3. [11]

    Evoli, P

    C. Evoli, P. Blasi, E. Amato and R. Aloisio,Signature of Energy Losses on the Cosmic Ray Electron Spectrum, Phys. Rev. Lett.125 (2020) 051101 [2007.01302]

    2020 arXiv

  4. [12]

    Fang, X.-J

    K. Fang, X.-J. Bi, S.-J. Lin and Q. Yuan,Klein–Nishina Effect and the Cosmic Ray Electron Spectrum, Chin. Phys. Lett.38 (2021) 039801 [2007.15601]

    2021 arXiv

  5. [13]

    Di Mauro, F

    M. Di Mauro, F. Donato and S. Manconi,Novel interpretation of the latest AMS-02 cosmic-ray electron spectrum, Phys. Rev. D104 (2021) 083012 [2010.13825]

    2021 arXiv

  6. [14]

    collaboration, The energy spectrum of cosmic-ray electrons at TeV energies, Phys

    H.E.S.S. collaboration, The energy spectrum of cosmic-ray electrons at TeV energies, Phys. Rev. Lett.101 (2008) 261104 [0811.3894]

    2008 arXiv

  7. [15]

    MAGIC collaboration, Measurement of the cosmic electron spectrum with the MAGIC telescopes, in32nd International Cosmic Ray Conference, vol. 6, pp. 47–50, 10, 2011, DOI [1110.4008]

    2011 arXiv

  8. [16]

    VERITAS collaboration, A Cosmic-ray Electron Spectrum with VERITAS, PoS ICRC2015 (2016) 411 [1508.06597]

    2016 arXiv

  9. [17]

    Aharonian, F.A

    F. Aharonian, F.A. Benkhali, J. Aschersleben, H. Ashkar, M. Backes, V.B. Martins et al., High-Statistics Measurement of the Cosmic-Ray Electron Spectrum with H.E.S.S., Phys. Rev. Lett.133 (2024) 221001 [2411.08189]

    2024

  10. [18]

    Recchia, S

    S. Recchia, S. Gabici, F.A. Aharonian and J. Vink,Local fading accelerator and the origin of TeV cosmic ray electrons, Phys. Rev. D99 (2019) 103022 [1811.07551]

    2019 arXiv

  11. [19]

    Fornieri, D

    O. Fornieri, D. Gaggero, D. Guberman, L. Brahimi, P. De La Torre Luque and A. Marcowith, Diffusive origin for the cosmic-ray spectral hardening reveals signatures of a nearby source in the leptons and protons data, Phys. Rev. D104 (2021) 103013 [2007.15321]. – 25 –

    2021 arXiv

  12. [20]

    Mertsch,Stochastic cosmic ray sources and the TeV break in the all-electron spectrum, JCAP 11 (2018) 045 [1809.05104]

    P. Mertsch,Stochastic cosmic ray sources and the TeV break in the all-electron spectrum, JCAP 11 (2018) 045 [1809.05104]

    2018 arXiv

  13. [21]

    Gaggero, L

    D. Gaggero, L. Maccione, G. Di Bernardo, C. Evoli and D. Grasso,Three-Dimensional Model of Cosmic-Ray Lepton Propagation Reproduces Data from the Alpha Magnetic Spectrometer on the International Space Station, Phys. Rev. Lett.111 (2013) 021102 [1304.6718]

    2013 arXiv

  14. [22]

    Boudaud, E.F

    M. Boudaud, E.F. Bueno, S. Caroff, Y. Genolini, V. Poulin, V. Poireau et al.,The pinching method for Galactic cosmic ray positrons: implications in the light of precision measurements, Astron. Astrophys.605 (2017) A17 [1612.03924]

    2017 arXiv

  15. [23]

    Boschini et al.,HelMod in the works: from direct observations to the local interstellar spectrum of cosmic-ray electrons, Astrophys

    M.J. Boschini et al.,HelMod in the works: from direct observations to the local interstellar spectrum of cosmic-ray electrons, Astrophys. J. 854 (2018) 94 [1801.04059]

    2018 arXiv

  16. [24]

    Z. Tian, W. Liu, B. Yang, X.-D. Fu, H.-B. Xu, Y.-h. Yao et al.,Electron and positron spectra in three-dimensional spatial-dependent propagation model, Chin. Phys. C 44 (2020) 085102 [1904.10663]

    2020 arXiv

  17. [25]

    Diesing and D

    R. Diesing and D. Caprioli,Spectrum of Electrons Accelerated in Supernova Remnants, Phys. Rev. Lett.123 (2019) 071101 [1905.07414]

    2019 arXiv

  18. [26]

    Kobayashi, Y

    T. Kobayashi, Y. Komori, K. Yoshida and J. Nishimura,The most likely sources of high energy cosmic-ray electrons in supernova remnants, Astrophys. J.601 (2004) 340 [astro-ph/0308470]

    2004 arXiv

  19. [27]

    Malyshev, I

    D. Malyshev, I. Cholis and J. Gelfand,Pulsars versus Dark Matter Interpretation of ATIC/PAMELA, Phys. Rev. D80 (2009) 063005 [0903.1310]

    2009 arXiv

  20. [28]

    Di Mauro et al.,Theoretical interpretation of Pass 8F ermi-LAT e+ + e− data, Astrophys

    M. Di Mauro et al.,Theoretical interpretation of Pass 8F ermi-LAT e+ + e− data, Astrophys. J. 845 (2017) 107 [1703.00460]

    2017 arXiv

  21. [29]

    Di Mauro, S

    M. Di Mauro, S. Manconi and F. Donato,Detection of aγ-ray halo around Geminga with the Fermi -LAT data and implications for the positron flux, Phys. Rev. D100 (2019) 123015 [1903.05647]

    2019 arXiv

  22. [30]

    Pohl and J.A

    M. Pohl and J.A. Esposito,Electron acceleration in SNR and diffuse gamma-rays above 1-GeV, Astrophys. J. 507 (1998) 327 [astro-ph/9806160]

    1998 arXiv

  23. [31]

    Blasi and E

    P. Blasi and E. Amato,Positrons from pulsar winds, Astrophys. Space Sci. Proc.(2011) 623 [1007.4745]

    2011 arXiv

  24. [32]

    Mertsch,Cosmic ray electrons and positrons from discrete stochastic sources, JCAP 02 (2011) 031 [1012.0805]

    P. Mertsch,Cosmic ray electrons and positrons from discrete stochastic sources, JCAP 02 (2011) 031 [1012.0805]

    2011 arXiv

  25. [33]

    Cholis, T

    I. Cholis, T. Karwal and M. Kamionkowski,Features in the Spectrum of Cosmic-Ray Positrons from Pulsars, Phys. Rev. D97 (2018) 123011 [1712.00011]

    2018 arXiv

  26. [34]

    Porter, G

    T.A. Porter, G. Jóhannesson and I.V. Moskalenko,Deciphering Residual Emissions: Time-dependent Models for the Nonthermal Interstellar Radiation from the Milky Way, ApJ 887 (2019) 250 [1909.02223]

    2019 arXiv

  27. [35]

    Evoli, E

    C. Evoli, E. Amato, P. Blasi and R. Aloisio,Galactic factories of cosmic-ray electrons and positrons, Phys. Rev. D103 (2021) 083010 [2010.11955]

    2021 arXiv

  28. [36]

    Evoli, E

    C. Evoli, E. Amato, P. Blasi and R. Aloisio,Stochastic nature of Galactic cosmic-ray sources, Phys. Rev. D104 (2021) 123029 [2111.01171]

    2021 arXiv

  29. [37]

    Cholis and I

    I. Cholis and I. Krommydas,Utilizing cosmic-ray positron and electron observations to probe the averaged properties of Milky Way pulsars, Phys. Rev. D105 (2022) 023015 [2111.05864]

    2022 arXiv

  30. [38]

    Orusa, S

    L. Orusa, S. Manconi, F. Donato and M. Di Mauro,Constraining positron emission from pulsar populations with AMS-02 data, JCAP 12 (2021) 014 [2107.06300]

    2021 arXiv

  31. [39]

    Asano, Y

    K. Asano, Y. Asaoka, Y. Akaike, N. Kawanaka, K. Kohri, H.M. Motz et al.,Monte Carlo Study of Electron and Positron Cosmic-Ray Propagation with the CALET Spectrum, Astrophys. J. 926 (2022) 5 [2111.09636]. – 26 –

    2022 arXiv

  32. [40]

    Marinos, T.A

    P.D. Marinos, T.A. Porter, G.P. Rowell, G. Jóhannesson and I.V. Moskalenko,On the Temporal Variability of the Galactic Multi-TeV Interstellar Emissions, Astrophys. J. 981 (2025) 93 [2411.03553]

    2025 arXiv

  33. [41]

    Germain, K

    M. Germain, K. Gregor, I. Murray and H. Larochelle,MADE: Masked Autoencoder for Distribution Estimation, arXiv e-prints (2015) arXiv:1502.03509 [1502.03509]

    2015 arXiv

  34. [42]

    Ahlers, P

    M. Ahlers, P. Mertsch and S. Sarkar,Cosmic ray acceleration in supernova remnants and the FERMI/PAMELA data, Phys. Rev. D80 (2009) [ 0909.4060]

    2009 arXiv

  35. [43]

    Vallée,The Spiral Arms and Interarm Separation of the Milky Way: An Updated Statistical Study, AJ 130 (2005) 569

    J.P. Vallée,The Spiral Arms and Interarm Separation of the Milky Way: An Updated Statistical Study, AJ 130 (2005) 569

    2005

  36. [44]

    Maurin, H.P

    D. Maurin, H.P. Dembinski, J. Gonzalez, I.C. Mariş and F. Melot,Cosmic-Ray Database Update: Ultra-High Energy, Ultra-Heavy, and Antinuclei Cosmic-Ray Data (CRDB v4.0), Universe 6 (2020) 102 [2005.14663]

    2020 arXiv

  37. [45]

    Tammann, W

    G.A. Tammann, W. Loeffler and A. Schroeder,The Galactic Supernova Rate, ApJS 92 (1994) 487

    1994

  38. [46]

    Rozwadowska, F

    K. Rozwadowska, F. Vissani and E. Cappellaro,On the rate of core collapse supernovae in the milky way, New A83 (2021) 101498 [2009.03438]

    2021 arXiv

  39. [47]

    Lagage and C.J

    P.O. Lagage and C.J. Cesarsky,The maximum energy of cosmic rays accelerated by supernova shocks., A&A 125 (1983) 249

    1983

  40. [48]

    Schwefer, P

    G. Schwefer, P. Mertsch and C. Wiebusch,Diffuse Emission of Galactic High-energy Neutrinos from a Global Fit of Cosmic Rays, Astrophys. J. 949 (2023) 16 [2211.15607]

    2023 arXiv

  41. [49]

    Bishop and H

    C.M. Bishop and H. Bishop,Deep Learning - Foundations and Concepts, Springer Cham (2023)

    2023

  42. [50]

    Goodfellow, Y

    I. Goodfellow, Y. Bengio and A. Courville,Deep Learning, MIT Press (2016)

    2016

  43. [51]

    B. Uria, I. Murray and H. Larochelle,RNADE: The real-valued neural autoregressive density-estimator, arXiv e-prints (2013) [ 1306.0186]

    2013 arXiv

  44. [52]

    HA WCcollaboration, Extended gamma-ray sources around pulsars constrain the origin of the positron flux at Earth, Science 358 (2017) 911 [1711.06223]

    2017 arXiv

  45. [53]

    Fang,Gamma-ray pulsar halos in the Galaxy, Front

    K. Fang,Gamma-ray pulsar halos in the Galaxy, Front. Astron. Space Sci.9 (2022) 1022100 [2209.13294]

    2022 arXiv

  46. [54]

    Cranmer, J

    K. Cranmer, J. Brehmer and G. Louppe,The frontier of simulation-based inference, Proceedings of the National Academy of Sciences117 (2020) 30055 [https://www.pnas.org/doi/pdf/10.1073/pnas.1912789117]. – 27 –

    2020 doi

Pith tools

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