Pith. sign in

REVIEW 3 major objections 5 minor 53 references

Deep learning inference with the Event Horizon Telescope III. Zingularity results from the 2017 observations and predictions for future array expansions

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read A neural network trained on simulated EHT images finds M87* in a retrograde magnetically arrested state, Sgr A* at high spin, and predicts the Africa Millimeter Telescope cuts non-Kerr errors threefold.

desk verdict First full-Stokes BANN inference on 2017 EHT data; Sgr A* is credible, M87* spin/Rhigh claims sit on training-grid boundaries and need reframing. read the letter →

arxiv 2506.13877 v1 pith:L4N7XBQS submitted 2025-06-16 astro-ph.HE astro-ph.IM

M. Janssen , C.-k. Chan , J. Davelaar , M. Wielgus This is my paper · ORCID
classification astro-ph.HEastro-ph.IM
keywords blackholeparameterinferenceEventHorizonTelescopeGRMHDsimulationsBayesianneuralnetworksM87*SgrA*verylongbaselineinterferometrygeneralrelativitytests
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 reports the first direct application of a Bayesian neural network to infer black-hole and accretion parameters from 2017 Event Horizon Telescope visibility measurements. The network, trained on a large library of ray-traced GRMHD simulations with simulated calibration errors, returns narrow posteriors that place M87* in a retrograde magnetically arrested accretion state with spin magnitude 0.5 to 0.94 and strong jet synchrotron emission, and Sgr A* at high spin (about 0.8 to 0.9), low inclination, and position angle 106 to 137 degrees, with no clear magnetic-state classification. The same machinery predicts that adding the Africa Millimeter Telescope will reduce non-Kerr parameter-inference errors by a factor of three, sharpening future tests of general relativity. If these inferences hold, they provide a fast, reproducible route from raw EHT visibilities to physical parameters and a quantitative guide for array design.

What carries the argument

Zingularity: a Bayesian artificial neural network that combines a residual network with variational fully connected layers, trained on synthetic EHT observations generated from ray-traced GRMHD simulations. The network maps full-Stokes visibility amplitudes and phases to posterior distributions over spin, electron-proton temperature ratio, MAD/SANE magnetic state, inclination, and position angle. Inference is repeated over 1000 bootstrap realizations of the observational data with simulated polarization leakage, gain and gain-curve errors, and thermal noise, so the reported posteriors include data systematics.

What would settle it

Apply the identical trained network to an independent EHT epoch for Sgr A* and to a later M87* campaign; if the inferred parameters shift outside the stated posteriors, the synthetic library is not representative of real data. A spectroscopic or polarimetric detection of a nonthermal electron distribution in M87*'s inner flow would similarly falsify the thermal temperature-ratio premise behind the jet-synchrotron interpretation.

Watch

Extended reading notes

Core claim

The paper's central claim is that, under its GRMHD model library, the 2017 EHT data are best explained by a retrograde magnetically arrested disk around M87* with spin magnitude between 0.5 and 0.94 and an electron-proton temperature ratio at the training-grid maximum, implying jet-dominated synchrotron emission; and by a high-spin, low-inclination, prograde flow around Sgr A* with position angle near 106 to 137 degrees, in a state beyond the standard MAD/SANE dichotomy. The authors present this as the first application of a Bayesian neural network trained on a synthetic GRMHD library directly to EHT visibilities, using full Stokes information and bootstrapping of known systematics. They also claim the same network predicts a threefold reduction in non-Kerr parameter-inference errors when the Africa Millimeter Telescope is added to the array. They interpret the M87* spin posterior as peaking at neighboring grid values and therefore argue the true spin is likely intermediate between the two, and they emphasize that the posteriors are insensitive to constant Faraday rotation and calibration gain biases.

Load-bearing premise

The result stands on the assumption that the finite set of simulated black-hole images used for training adequately represents the real 2017 EHT data, including its calibration errors, polarization leakage, and Faraday rotation; if the real data contain physics absent from the simulations, the inferred posteriors would not follow.

Editorial extensions

If this is right

  • If M87* is truly in a retrograde magnetically arrested state, its powerful jet and counter-rotation fit a merger history, and the inferred parameters satisfy jet-power constraints measured on larger scales.
  • If Sgr A* has high spin and a spin axis nearly aligned with our line of sight, the polarization-loop direction from GRAVITY favors the inclination solution that makes the accretion flow rotate clockwise on the sky.
  • Because the network is insensitive to the constant Faraday rotation measure, the posteriors do not require assumptions about where the rotation measure originates, removing a known systematic in previous GRMHD scoring.
  • Adding the Africa Millimeter Telescope should reduce non-Kerr parameter-inference errors by roughly a factor of three, primarily through improved northeast-southwest resolution and short-baseline calibration tracks.
  • Producing new GRMHD simulations at the inferred interpolated parameters would allow direct model-data comparisons of accretion rate, jet power, and broadband spectral energy distributions.

Reading between the lines

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

  • Beyond the paper: the method treats the synthetic GRMHD library as the prior, so the reported posteriors are conditional on that library; retraining on a library that includes nonthermal electron distributions, tilted disks, or different magnetic field polarities could shift the inferred parameters.
  • Beyond the paper: the demonstrated insensitivity to constant Faraday rotation could be exported to other polarized VLBI sources, letting the network separate foreground screen rotation from intrinsic source structure without explicit rotation-measure fitting.
  • Beyond the paper: the failure to train on Kerr-Newman models points to a spin-charge degeneracy in current baseline coverage; a network that recovers both spin and charge would provide a direct, quantitative test of whether future arrays can break the degeneracy.
  • Beyond the paper: a cheap validation would be to apply the same trained networks to the 2018 or 2021 EHT epochs; stable posteriors across epochs would strengthen the claim that the 2017 result is not an artifact of that particular observing run.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper applies Bayesian neural networks (the Zingularity framework) trained on a large library of ray-traced GRMHD synthetic EHT observations to the 2017 EHT data for M87* and Sgr A*. It uses 1000 bootstrapped realizations of the data with simulated gain errors, polarization leakage, thermal noise, and Faraday rotation, and reports posteriors for spin, electron heating ratio Rhigh, magnetic state, inclination, and position angle. The main results are that M87* is best described by a retrograde MAD accretion flow with Rhigh near the training maximum of 160 and spin between -0.5 and -0.94, while Sgr A* prefers high spin (~0.8-0.9), low inclination, position angle 106-137 degrees, and an inconclusive MAD/SANE state. The paper also predicts that the proposed Africa Millimeter Telescope (AMT) extension will reduce parameter inference errors by a factor of three for non-Kerr models.

Significance. If the M87* inference were clean, this would be a significant methodological demonstration: full-Stokes Bayesian neural network inference on EHT visibilities, with explicit bootstrapping of known systematics and a substantially larger synthetic library than earlier work. The paper has real strengths: reproducible data reduction scripts, public availability of the training data and code, careful validation diagnostics, and unusually candid acknowledgment of the limitations of the discrete GRMHD grid. The Sgr A* results are plausible and largely interior to the training grid, and the AMT forecasting exercise is useful. However, the M87* headline parameters sit at or between discrete training grid values, and the paper itself concedes these may be artifacts; as a result the M87* half of the central claim is not yet established at the level claimed in the abstract.

major comments (3)
  1. [Section 4, Fig. 4, and Section 5] The M87* Rhigh posterior peaks at Rhigh = 159.91 with the training maximum at 160, and the text concedes that a model with an even higher Rhigh might describe the data better. A posterior peaking at the edge of the training range cannot support the abstract's statement that M87* is 'best described by' Rhigh=160; at most it supports a lower limit under the current library. This should be fixed by extending the Rhigh grid or by explicitly reporting the result as censored at the training boundary.
  2. [Section 4, Fig. 4, and Section 5] The M87* spin posterior is bimodal at a* = -0.5 and -0.94, which Section 5 identifies as two neighboring values in the GRMHD training grid. The abstract's 'spin between 0.5 and 0.94' is therefore the span of two discrete labels, not a measured credible interval. The paper needs a refined spin grid to test whether the bimodality persists, or the claim must be downgraded to 'the data are consistent with retrograde spins in this range, with the posterior location currently driven by grid discreteness.'
  3. [Section 3, Fig. 3, and Section 4] The claim that the inference is 'without being impacted by the unknown foreground Faraday screens and data calibration biases' is supported by internal validation on the synthetic library and by the RM de-rotation test, but that test only varies the constant RM component. The paper states that applying a time-variable RM changes the results because the network uses internal time-variability of Q and U phases as discriminating features. Since the training library is the only source of that variability, the posterior widths for Sgr A* inclination and spin may be under-calibrated if real Faraday variability differs from the simulated one. A coverage test on held-out simulations with perturbed Faraday and calibration parameters would directly address this concern.
minor comments (5)
  1. [Title] The header and abstract run 'Zingularityresults' together; there should be a space between 'Zingularity' and 'results'.
  2. [Section 3] The sentence 'For Kerr Sgr A∗, we two equally viable models' is missing a verb and should read 'we have two equally viable models'.
  3. [Section 4] The sentence 'The low Rhigh∼ 14 value corresponds to a SANE MAD = 0.36+0.25−0.20' is unclear; it should specify that this is the inferred value associated with a particular magnetic-state classification, not a definition of Rhigh.
  4. [Section 5] The phrase 'The small 10◦ difference in inclination angle' does not specify which two inferences are being compared; please clarify whether it refers to the two fiducial Sgr A* networks or the fiducial versus dilaton network.
  5. [Sections 4 and 5] The discrete parameter grid (spin values, Rhigh values, ilos and PA step sizes) is not fully stated in this paper; it should be listed explicitly so that boundary and grid-discreteness effects can be assessed without consulting the companion papers.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation found: the BANN is a learned map from synthetic visibilities to GRMHD labels, and the headline M87* grid-edge behavior is a discretization limitation, not a circular step.

full rationale

The paper's inference chain is: (1) generate synthetic EHT observations from GRMHD simulations with known labels; (2) train Bayesian neural networks to map visibility data to those labels; (3) apply the trained networks to calibrated 2017 EHT data with bootstrapped systematics. None of these steps defines the target parameters in terms of the network output, and no fitted parameter is renamed as a prediction. The M87* posterior peaking at Rhigh=160, the training maximum, and at adjacent spin grid values is explicitly acknowledged in Section 5: 'The training data had a maximum Rhigh value of 160, so it might be that a model with an even higher Rhigh would describe the data better' and 'The probability density likely peaks at those particular values because they are two neighboring values in our GRMHD training data grid space.' This is a grid-resolution and extrapolation caveat that limits the strength of the M87* claim, but it is not circular: the posterior still comes from real data evaluated by a network trained on labeled simulations. The citations to Janssen et al. 2025a,b are methodological self-citations describing the synthetic library and network validation; they are not invoked as an unverified uniqueness theorem, and the paper points to code and validation diagnostics rather than relying on authority alone. The AMT factor-of-three statement is a forward simulation comparison using the same validated pipeline, not a restatement of an input. Overall, no equation-level or definitional circularity is present.

Assumptions & free parameters 8 free parameters · 6 assumptions · 0 invented entities

The central inference rests on a closed loop between the GRMHD library, the synthetic data generator, and the BANN. The free parameters shown here are either the physical inference targets or the grid and hyperparameter choices that shape the posteriors. The axioms are the modeling assumptions needed for the synthetic-to-real transfer. No new physical entities are postulated.

free parameters (8)
  • M87* spin a* = -0.5 to -0.94 (retrograde)
    Inferred by the BANN from the April 11 2017 data; central result of the paper.
  • M87* Rhigh electron heating parameter = 160 (training-grid maximum)
    Posterior peaks at the boundary of the training library; the authors note an even higher value might fit better.
  • Sgr A* spin a* = 0.8 to 0.9 (prograde)
    Inferred by the network from April 7 2017 data; central result.
  • Sgr A* Rhigh electron heating parameter = about 14
    Network-inferred value corresponding to hot electrons; used to argue against a powerful jet.
  • Sgr A* inclination ilos = 29 deg or 162 deg (18 deg corrected)
    Two fiducial networks give the two symmetric solutions; the authors prefer the >90 deg solution from Q-U loop constraints.
  • Sgr A* position angle PA = 106 to 137 deg
    The inferred PA values bracket the training grid values 90, 120, and 150 deg, so the range partly reflects the discretization.
  • GRMHD parameter grid discretization = spin grid, Rhigh up to 160, ilos in 20 deg steps, PA in 90/120/150
    The paper admits the M87* posterior peaks at two neighboring grid values, so the grid itself shapes the quoted intervals.
  • BANN architecture and training hyperparameters = selected via parameter surveys in companion paper
    Network architecture, activation, optimizer, epochs, and regularization were chosen by survey rather than derived, and they affect the quoted validation errors.
assumptions (6)
  • domain assumption GRMHD ray-traced simulations accurately describe the horizon-scale emission of Sgr A* and M87*.
    Section 5 states GRMHD simulations are currently the best self-consistent models of the horizon-scale emission; this is a modeling premise, not independently verified.
  • domain assumption Synthetic EHT training data, including simulated gain errors, polarization leakage, and thermal noise, are representative of real 2017 EHT systematics.
    Sections 2 and 3 say training sets are 'akin to the CASA data' and rely on simulated corruption; no real-data transfer test is shown.
  • domain assumption The BANN can validly interpolate between the discrete GRMHD parameter grid values.
    Section 4 says the BANNs are interpolating between training values; Section 5 shows grid discretization artifacts in M87* and PA, so interpolation is imperfect.
  • domain assumption A constant external Faraday rotation measure does not affect the inference.
    Section 5 shows identical posteriors when de-rotating Q,U by 50% or 100% of the RM, but also says a time-variable RM changes results, so robustness is limited.
  • domain assumption Kerr spacetime with fixed black hole masses for Sgr A* and M87*.
    Section 1 fixes masses to 4.14e6 and 6.2e9 solar masses and uses static Kerr metric for standard models.
  • domain assumption Thermal electron distribution, pure hydrogen gas, and no pair production or radiative cooling in the training simulations.
    Section 5 lists these as explicit limitations of the GRMHD library.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Deep learning inference with the Event Horizon Telescope III. Zingularity results from the 2017 observations and predictions for future array expansions." pith.science (2026). https://pith.science/paper/L4N7XBQS

@misc{pith2026250613877,
  author       = {Pith},
  title        = {Pith review of: Deep learning inference with the Event Horizon Telescope III. Zingularity results from the 2017 observations and predictions for future array expansions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/L4N7XBQS}},
  note         = {Machine review of arXiv:2506.13877}
}
abstract

(abridged) In the first two papers of this publication series, we present a comprehensive library of synthetic EHT observations and used this library to train and validate Bayesian neural networks for the parameter inference of accreting supermassive black hole systems. The considered models are ray-traced GRMHD simulations of Sgr A* and M87*. In this work, we infer the best-fitting accretion and black hole parameters from 2017 EHT data and predict improvements that will come with future upgrades of the array. Compared to previous EHT analyses, we considered a substantially larger synthetic data library and the most complete set of information from the observational data. We made use of the Bayesian nature of the trained neural networks and apply bootstrapping of known systematics in the observational data to obtain parameter posteriors. Within a wide GRMHD parameter space, we find M87* to be best described by a spin between 0.5 and 0.94 with a retrograde MAD accretion flow and strong synchrotron emission from the jet. Sgr A* has a high spin of $\sim$ 0.8 $-$ 0.9 and a prograde accretion flow beyond the standard MAD/SANE models with a comparatively weak jet emission, seen at a $\sim$ 20$^\circ$ $-$ 40$^\circ$ inclination and $\sim$ 106$^\circ$ $-$ 137$^\circ$ position angle. While previous EHT analyses could rule out specific regions in the model parameter space considered here, we are able to obtain narrow parameter posteriors with our Zingularity framework without being impacted by the unknown foreground Faraday screens and data calibration biases. We further demonstrate that the AMT extension to the EHT will reduce parameter inference errors by a factor of three for non-Kerr models, enabling more robust tests of general relativity. It will be instructive to produce new GRMHD models with the inferred interpolated parameters to study their accretion rate plus jet power.

Figures

Figures reproduced from arXiv: 2506.13877 by the authors.

Figure 1
Figure 1. Correlated flux densities in Jansky (Jy) and visibility phases in degrees (deg) with standard deviation error bands computed from 1000 bootstrapping realization of the 2017 April 11 M87∗ (left panels) and April 7 Sgr A∗ (right panels) observational EHT data. The measurements are plotted as a function of baseline length in units of the λ 1.3 mm observing wavelength. The displayed Stokes parameters show the total inte… view at source ↗
Figure 2
Figure 2. Visualization of training datasets with parameters close to our best-fitting BANN inferences. The top row displays example total intensity ray-traced ground-truth model images on logarithmic scales with varying dynamic ranges. Top left corners show the type of model, top right corners show the model parameters: spin a∗ = s, Rhigh = r, and ilos = l parameters are listed in a shorthand notation as as , Rr , and il . T… view at source ↗
Figure 3
Figure 3. Zingularity performance diagnostics for various neural network training runs. The validation error is computed from normalized labels of validation data not seen by the network during training. The mean absolute error (MAE) is computed as the average of all validation samples for normalized regression labels. The classification error (Class. error) is defined as one minus the network’s accuracy, i.e., the fraction o… view at source ↗
Figures from the paper (2 more)
Figure 5
Figure 5. Figure 5: shows the results from applying our two fiducial Sgr A∗ BANNs to the April 7 observational data. The data are inconclusive concerning the magnetic state of the accretion flow. MAD models describe the polarization quantities well, while SANE models are less problematic …
Figure 6
Figure 6. Figure 6: Same as [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 28 canonical work pages

  1. [1]

    2025, A&A, 696, A10

    Antonopoulou, E., Loules, A., & Nathanail, A. 2025, A&A, 696, A10

    2025

  2. [2]

    & Nathanail, A

    Antonopoulou, E. & Nathanail, A. 2024, A&A, 690, A240

    work page 2024

  3. [3]

    E., et al

    Backes, M., Müller, C., Conway, J. E., et al. 2016, in The 4th Annual Conference on High Energy Astrophysics in Southern Africa (HEASA 2016), 29

    2016

  4. [4]

    2021, ApJ, 917, 8

    Ball, D., Özel, F., Christian, P., Chan, C.-K., & Psaltis, D. 2021, ApJ, 917, 8

    work page 2021

  5. [5]

    & V olonteri, M

    Berti, E. & V olonteri, M. 2008, ApJ, 684, 822

    work page 2008

  6. [6]

    2017, MNRAS, 464, 143

    Blecher, T., Deane, R., Bernardi, G., & Smirnov, O. 2017, MNRAS, 464, 143

    2017

  7. [7]

    D., et al

    Chavez, E., Issaoun, S., Johnson, M. D., et al. 2024, ApJ, 974, 116

    work page 2024

  8. [8]

    A., Donahue, M., O’Dea, C

    Daly, R. A., Donahue, M., O’Dea, C. P., et al. 2024, MNRAS, 527, 428

    work page 2024

Show all 53 references
  1. [9]

    2019, in Bulletin of the American Astronomical Society, V ol

    Doeleman, S., Blackburn, L., Dexter, J., et al. 2019, in Bulletin of the American Astronomical Society, V ol. 51, 256

    2019

  2. [10]

    S., Barrett, J., Blackburn, L., et al

    Doeleman, S. S., Barrett, J., Blackburn, L., et al. 2023, Galaxies, 11, 107

    2023

  3. [11]

    A., Zajacek, M., et al

    Eckart, A., Tursunov, A. A., Zajacek, M., et al. 2018, in Accretion Processes in Cosmic Sources - II, 48 Event Horizon Telescope Collaboration, Akiyama, K., Alberdi, A., et al. 2024, ApJ, 964, L26 Event Horizon Telescope Collaboration, Akiyama, K., Alberdi, A., et al. 2023, Ap...

    2018

  4. [12]

    C., Özel, F., & Psaltis, D

    Faggert, J. C., Özel, F., & Psaltis, D. 2025, ApJ, 981, 209

    2025

  5. [13]

    2016, The Journal of Open Source Software, 1, 24

    Foreman-Mackey, D. 2016, The Journal of Open Source Software, 1, 24

    2016

  6. [14]

    2025, ApJ, 978, 148 García, A., Galtsov, D., & Kechkin, O

    Galishnikova, A., Philippov, A., Quataert, E., Chatterjee, K., & Liska, M. 2025, ApJ, 978, 148 García, A., Galtsov, D., & Kechkin, O. 1995, Phys. Rev. Lett., 74, 1276

    2025

  7. [15]

    2021, ApJ, 910, L14

    Goddi, C., Martí-Vidal, I., Messias, H., & et al. 2021, ApJ, 910, L14

    2021

  8. [16]

    E., Younsi, Z., et al

    Gold, R., Broderick, A. E., Younsi, Z., et al. 2020, ApJ, 897, 148 Gravity Collaboration, Abuter, R., Aimar, N., et al. 2023, A&A, 677, L10 Gravity Collaboration, Abuter, R., Amorim, A., et al. 2018, A&A, 618, L10 GRA VITY Collaboration, Bauböck, M., Dexter, J., et al. 2020, A...

    2020

  9. [17]

    2015, arXiv e-prints, arXiv:1512.03385

    He, K., Zhang, X., Ren, S., & Sun, J. 2015, arXiv e-prints, arXiv:1512.03385

    2015 arXiv

  10. [18]

    H., et al

    Helmi, A., Babusiaux, C., Koppelman, H. H., et al. 2018, Nature, 563, 85

    2018

  11. [19]

    D., Blackburn, L., et al

    Issaoun, S., Johnson, M. D., Blackburn, L., et al. 2019, ApJ, 871, 30

    2019

  12. [20]

    2018, in 14th European VLBI Network Symposium & Users Meeting (EVN 2018), 80

    Janssen, M., Goddi, C., Falcke, H., et al. 2018, in 14th European VLBI Network Symposium & Users Meeting (EVN 2018), 80

    2018

  13. [21]

    M., et al

    Janssen, M., Goddi, C., van Bemmel, I. M., et al. 2019, A&A, 626, A75

    2019

  14. [22]

    F., & Wagner, J

    Janssen, M., Radcliffe, J. F., & Wagner, J. 2022, Universe, 8, 527

    2022

  15. [23]

    V ., Prather, B

    Joshi, A. V ., Prather, B. S., Chan, C.-k., Wielgus, M., & Gammie, C. F. 2024, ApJ, 972, 135

    2024

  16. [24]

    Kerr, R. P. 1963, Phys. Rev. Lett., 11, 237

    1963

  17. [25]

    M., Dai, L., & Tchekhovskoy, A

    Kwan, T. M., Dai, L., & Tchekhovskoy, A. 2023, ApJ, 946, L42 La Bella, N., Issaoun, S., Roelofs, F., Fromm, C., & Falcke, H. 2023, A&A, 672, A16

    2023

  18. [26]

    A., Bouman, K

    Levis, A., Chael, A. A., Bouman, K. L., Wielgus, M., & Srinivasan, P. P. 2024, Nature Astronomy, 8, 765 Article number, page 8 of 9 M. Janssen et al.: Deep learning inference with the Event Horizon Telescope

    2024

  19. [27]

    2023, MNRAS, 520, 1271

    Lin, X., Li, Y .-P., & Yuan, F. 2023, MNRAS, 520, 1271

    2023

  20. [28]

    2018, MNRAS, 474, L81

    Liska, M., Hesp, C., Tchekhovskoy, A., et al. 2018, MNRAS, 474, L81

    2018

  21. [29]

    2024, ApJ, 960, 82

    Lowell, B., Jacquemin-Ide, J., Tchekhovskoy, A., & Duncan, A. 2024, ApJ, 960, 82

    2024

  22. [30]

    2020, MNRAS, 497, 2385

    Matsumoto, T., Chan, C.-H., & Piran, T. 2020, MNRAS, 497, 2385

    2020

  23. [31]

    M., et al

    Mizuno, Y ., Younsi, Z., Fromm, C. M., et al. 2018, Nature Astronomy, 2, 585 Mo´scibrodzka, M. 2025, ApJ, 981, 145 Mo´scibrodzka, M. & Falcke, H. 2013, A&A, 559, L3 Mo´scibrodzka, M., Falcke, H., & Shiokawa, H. 2016, A&A, 586, A38 Mo´scibrodzka, M., Gammie, C. F., Dolence, J. ...

    2018

  24. [32]

    M., Phinney, E

    Murchikova, E. M., Phinney, E. S., Pancoast, A., & Blandford, R. D. 2019, Nature, 570, 83

    2019

  25. [33]

    2022, MNRAS, 511, 3795

    Narayan, R., Chael, A., Chatterjee, K., Ricarte, A., & Curd, B. 2022, MNRAS, 511, 3795

    2022

  26. [34]

    2022, MNRAS, 512, 490

    Natarajan, I., Deane, R., Martí-Vidal, I., et al. 2022, MNRAS, 512, 490

    2022

  27. [35]

    2019, ApJ, 880, L26

    Nemmen, R. 2019, ApJ, 880, L26

    2019

  28. [36]

    T., Couch, E., Chinnapared, K., et al

    Newman, E. T., Couch, E., Chinnapared, K., et al. 1965, Journal of Mathematical Physics, 6, 918 Özel, F., Psaltis, D., & Younsi, Z. 2022, ApJ, 941, 88

    1965

  29. [37]

    S., Dexter, J., Moscibrodzka, M., et al

    Prather, B. S., Dexter, J., Moscibrodzka, M., et al. 2023, ApJ, 950, 35

    2023

  30. [38]

    2023, MNRAS, 520, 4867

    Qiu, R., Ricarte, A., Narayan, R., et al. 2023, MNRAS, 520, 4867

    2023

  31. [39]

    I., Malkan, M., & Vestergaard, M

    Raimundo, S. I., Malkan, M., & Vestergaard, M. 2023, Nature Astronomy, 7, 463

    2023

  32. [40]

    Ramachandran, P., Zoph, B., & Le, Q. V . 2017, arXiv e-prints, arXiv:1710.05941

    2017 arXiv

  33. [41]

    M., White, C

    Ressler, S. M., White, C. J., & Quataert, E. 2023, MNRAS, 521, 4277

    2023

  34. [42]

    2021, MNRAS, 505, 523

    Ricarte, A., Qiu, R., & Narayan, R. 2021, MNRAS, 505, 523

    2021

  35. [43]

    2023, Galaxies, 11, 6 Röder, J., Cruz-Osorio, A., Fromm, C

    Ricarte, A., Tiede, P., Emami, R., Tamar, A., & Natarajan, P. 2023, Galaxies, 11, 6 Röder, J., Cruz-Osorio, A., Fromm, C. M., et al. 2023, A&A, 671, A143

    2023

  36. [44]

    2020, A&A, 636, A5

    Roelofs, F., Janssen, M., Natarajan, I., et al. 2020, A&A, 636, A5

    2020

  37. [45]

    J., Yusef-Zadeh, F., Wardle, M., et al

    Royster, M. J., Yusef-Zadeh, F., Wardle, M., et al. 2019, ApJ, 872, 2

    2019

  38. [46]

    Salas, L. D. S., Liska, M. T. P., Markoff, S. B., et al. 2025, MNRAS, 538, 698

    2025

  39. [47]

    Sarkar, K. C. 2024, A&A Rev., 32, 1

    2024

  40. [48]

    W., & Stone, J

    Sharma, P., Quataert, E., Hammett, G. W., & Stone, J. M. 2007, ApJ, 667, 714 V olonteri, M., Sikora, M., & Lasota, J.-P. 2007, ApJ, 667, 704

    2007

  41. [49]

    & Zhang, B

    Wang, Y . & Zhang, B. 2024, Nature Astronomy, 8, 1592

    2024

  42. [50]

    2024, A&A, 682, A97

    Wielgus, M., Issaoun, S., Martí-Vidal, I., et al. 2024, A&A, 682, A97

    2024

  43. [51]

    2023, ApJ, 942, 47

    Younsi, Z., Psaltis, D., & Özel, F. 2023, ApJ, 942, 47

    2023

  44. [52]

    2002, A&A, 383, 854

    Yuan, F., Markoff, S., & Falcke, H. 2002, A&A, 383, 854

    2002

  45. [53]

    2020, MNRAS, 499, 3909 Article number, page 9 of 9

    Yusef-Zadeh, F., Royster, M., Wardle, M., et al. 2020, MNRAS, 499, 3909 Article number, page 9 of 9

    2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.