Pith. sign in

REVIEW 4 major objections 3 minor 1 cited by

X-ray and radio polarimetry of the neutron star low mass X-ray binary GX 13+1

T0 review · 4 major / 3 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read Joint X-ray and radio polarimetry reads GX 13+1's accretion geometry as a soft disk plus a differently aligned boundary layer.

desk verdict First IXPE polarization of GX 13+1 with a plausible disk plus boundary-layer geometry, but the supplied full text is a different paper, leaving the load-bearing fit unverifiable. read the letter →

arxiv 2508.05763 v1 pith:GGVK7XFP submitted 2025-08-07 astro-ph.HE

Unnati Kashyap , Thomas J. Maccarone , Eliot C. Pattie , Mason Ng , Swati Ravi , Herman L. Marshall This is my paper
classification astro-ph.HE
keywords X-raypolarimetryradioneutronstarlow-massbinaryGX13+1accretiondiskboundarylayerdipsIXPE
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 uses simultaneous IXPE X-ray polarimetry, VLA radio polarimetry, and NICER spectral monitoring to ask where the polarized X-rays from the neutron-star low-mass X-ray binary GX 13+1 are produced and what that reveals about its accretion geometry. It reports that the source was in parts of the neutron-star Z state during the observations, showed X-ray dips, and gave hints of a polarization swing between dip and non-dip intervals. The central interpretive claim is a two-component spectropolarimetric picture: a soft accretion-disk component and a harder blackbody from the boundary layer or spreading layer, with different polarization properties. If this decomposition is right, X-ray polarization becomes a direct geometric probe of the boundary/spreading-layer orientation relative to the disk, a region that is otherwise hard to resolve in low-mass X-ray binaries.

What carries the argument

The carrying mechanism is a joint X-ray spectro-polarimetric decomposition: fitting the IXPE energy-resolved polarization data with two emission components, an accretion-disk continuum for the soft band and a blackbody for the harder boundary/spreading layer, so that each component's polarization fraction and angle are separated rather than averaged. The dip/non-dip division of the light curve provides a second handle, using X-ray dips to modulate the relative contribution of the components; the VLA radio polarization adds an independent orientation constraint on the accretion flow or outflow.

What would settle it

Re-fit the existing IXPE observation with fixed dip-phase bins: if the recovered polarization angle of the hard blackbody component is statistically identical to that of the soft disk, or if the dip/non-dip swing vanishes whenever the data are rebinned by dip phase, the inferred two-component geometry is not supported. A single-component Comptonized model that fits the same spectropolarimetric data without a separate boundary-layer blackbody would also falsify the specific disk-plus-boundary-layer decomposition.

Watch

Extended reading notes

Core claim

On its own terms, the paper claims that GX 13+1's X-ray polarization is not a single monolithic signal but a superposition: softer emission arising from the accretion disk and harder emission from a boundary layer or spreading layer around the neutron star, with the two components contributing different polarization fractions and angles. The strongest evidence quoted is the X-ray spectro-polarimetric fit, which prefers this disk-plus-blackbody decomposition; the dip/non-dip comparison shows only hints of polarization swings. The paper also combines X-ray and radio polarization findings to constrain the three-dimensional geometry of the binary. The claim matters because boundary and spreading

Load-bearing premise

The geometry follows only if the IXPE spectra cleanly separate a soft disk component from a hard boundary/spreading-layer blackbody with genuinely different polarization angles, and if co-adding dip and non-dip intervals does not mix states whose individual polarizations differ.

Editorial extensions

If this is right

  • If the disk-plus-boundary-layer decomposition is correct, IXPE-style spectropolarimetry can map the relative orientation of boundary/spreading layers and accretion disks in other low-mass X-ray binaries without spatial resolution.
  • A real dip/non-dip polarization swing would mean the spectral changes during dips are accompanied by a geometric reweighting of the polarized components, linking the dipping absorber to the inner flow geometry.
  • Combining radio and X-ray polarization angles offers a way to test whether the radio-emitting region shares the disk plane or is aligned with a jet or outflow axis.
  • The NICER-based Z-state classification anchors the polarization behavior to a known spectral state, so future state-resolved polarization observations can be compared directly.

Reading between the lines

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

  • Editorial caveat: the full text attached to this arXiv record is an unrelated robot-control paper, so the claims summarized here rest on the abstract alone; the numerical polarization results and fit details could not be checked in the supplied body.
  • If the dip/non-dip swing is confirmed with more exposure, dipping low-mass X-ray binaries could be used as natural polarization modulators, with the dip phase isolating contributions from different radii.
  • The same two-component spectropolarimetric fit could be applied to non-dipping sources to test whether boundary-layer/disk misalignment is a general feature of neutron-star accretion or specific to strongly dipping systems.
  • Simultaneous radio observations at other frequencies, or radio imaging, could test whether the VLA-measured polarization angle tracks a jet axis rather than the disk plane.
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 / 3 minor

Summary. The submission is titled 'X-ray and radio polarimetry of the neutron star low mass X-ray binary GX 13+1' and its abstract reports an IXPE/VLA/NICER study of the Z-source GX 13+1. The abstract claims that the source exhibits X-ray dips and 'hints of polarization swings' between dip and non-dip states, and that X-ray spectro-polarimetry suggests a two-component geometry: a soft accretion disk and a harder boundary/spreading layer with distinct polarization properties. However, the full text supplied is not the polarimetry paper: it is a robotics paper, 'GPU-Accelerated Barrier-Rate Guided MPPI Control for Tractor-Trailer Systems' (arXiv:2508.05773), containing no astronomical observations, data analysis, figures, or tables relevant to GX 13+1. No methods, spectral or polarimetric fits, uncertainties, or model comparisons are present for the abstract's claims. The only substantive evidence available for evaluation is the abstract itself.

Significance. If the abstract's conclusions were supported by a proper analysis, the paper would be of interest to the X-ray polarimetry and LMXB community: IXPE constraints on the relative orientation of an accretion disk and a boundary/spreading layer in a Z-source would be a valuable addition. The abstract's cautious phrasing ('hints', 'suggest') is appropriate given the evidently limited statistical power. However, the significance cannot be assessed because the body text is unrelated to the claimed study. The paper as submitted provides no derivations, no fits, no significance levels, no reproducibility artifacts, and no falsifiable quantitative predictions beyond qualitative statements. The scientific contribution is therefore unverifiable in this form.

major comments (4)
  1. [Full text (arXiv:2508.05773)] The full text is a robotics/control paper on tractor-trailer path planning, not the X-ray/radio polarimetry study described in the title and abstract. None of the central claims—the IXPE and VLA observations, the NICER hardness-intensity diagram, the dip/non-dip light curves, or the spectro-polarimetric decomposition into disk and boundary/spreading layer—are supported by any methods, equations, tables, or figures in the submitted manuscript. This is a load-bearing defect: the reader cannot check the fit, the error bars, the model comparison, or the geometry inference. The manuscript must be replaced with the actual polarimetry paper before any further evaluation.
  2. [Abstract, polarization swings] The abstract states only 'hints of polarization swings between the dip and non-dip states.' No significance, test statistic, or confidence interval is given. The phrase 'hints' is consistent with a noise fluctuation, yet the later sentence 'The X-ray spectro-polarimetry results suggest a source geometry...' treats the two-component interpretation as if it were established. The central geometry conclusion depends on this marginal signal and on the identifiability of the two-component decomposition; neither is quantified or justified in any retrievable part of the submission.
  3. [Abstract, two-component decomposition] The claim that the data require a soft disk component plus a blackbody/boundary-layer component with distinct polarization angles and fractions is a model-dependent inference. No spectral model comparison, goodness-of-fit values, or posterior uncertainties are reported. For IXPE data on a Z-source, where each component's polarization is expected to be small, the Stokes spectra of a two-component model can be degenerate with a single energy-dependent polarization model. The submission provides no evidence that the decomposition is identifiable, so the geometric conclusion is not supported.
  4. [Abstract, dip/non-dip state selection] The abstract reports co-added dip and non-dip intervals but gives no details on how these intervals were defined or whether the spectral state was stable within each interval. Absorption dips in LMXBs are often associated with changes in absorption column and possibly spectral shape; co-adding variable intervals can produce an apparent polarization angle change even without a change in intrinsic geometry. The manuscript contains no analysis controlling for this effect, so the 'polarization swing' remains an unquantified systematic risk.
minor comments (3)
  1. [Title/Abstract] The title and abstract are for an astrophysics paper while the full text is for a different paper in robotics. At minimum, the submission must be corrected so that the title, abstract, and body correspond to the same work.
  2. [Abstract] The abstract mentions 'moderate changes in the hardness intensity diagram' but provides no figure, quantitative hardness values, or observing dates. If the actual analysis were included, these would be needed for reproducibility.
  3. [Full text] The full text contains no references to IXPE, VLA, NICER, GX 13+1, or any astronomical data analysis. The reference list and notation are entirely from the robotics domain, making it impossible to cross-check any claim in the abstract.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the geometry claim is standard inverse inference from IXPE spectro-polarimetry, not a derivation from its own conclusion.

full rationale

The abstract's load-bearing claim is that X-ray spectro-polarimetry suggests an accretion disk plus a blackbody representing the boundary/spreading layer. This is an interpretation of measured Stokes parameters using conventional spectral and polarization models; it is inverse inference from data, not a prediction derived from a fitted parameter renamed as an output. There is no equation in the supplied abstract that defines the disk or blackbody component in terms of the inferred geometry, and no fitted quantity is presented as an independent prediction. The 'hints of polarization swings' is explicitly hedged language and is not used to force the geometry conclusion. No self-citation, no imported uniqueness theorem, and no ansatz smuggled via prior work appear in the abstract. The supplied full text is an unrelated robotics-control manuscript, so the statistical details, model comparisons, and error bars of the IXPE analysis cannot be checked from the provided material; however, unverifiability due to missing full text is a correctness/evidence concern, not circularity under the stated criteria. The only circular-adjacent issue is model dependence of the inferred polarization angles on the chosen two-component decomposition, but that is ordinary model dependence rather than a self-referential reduction.

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

The central claim is an inference from fitted spectral and polarization models rather than a parameter-free derivation. The free parameters are standard continuum and polarization fit parameters, whose values are not stated in the abstract. The axioms are domain assumptions about the two-component decomposition and the dip state selection. No new entities are invented; the boundary layer and spreading layer are established concepts from prior literature.

free parameters (3)
  • Two-component continuum fit parameters (N_H, disk kT, BL blackbody kT, normalizations) = not stated in abstract
    Standard absorbed disk plus blackbody fit to IXPE/NICER spectra; values carry the spectral decomposition on which the geometry claim rests.
  • Polarization fraction and angle per spectral component = not stated in abstract
    The spectropolarimetric fit outputs; any dip versus non-dip difference in these values is the reported 'hints of polarization swings'.
  • Geometry parameters derived from polarization angles (relative orientation of disk and BL/SL) = not stated in abstract
    The inferred geometry is a transformation of fitted polarization angles; degeneracies between component angles are a known risk in two-component IXPE decompositions.
assumptions (3)
  • domain assumption The IXPE Stokes spectra are separable into two components: a softer accretion disk and a harder boundary layer or spreading layer blackbody.
    The central geometry claim is only as strong as this decomposition; entered at the abstract's statement that spectro-polarimetry 'suggests a source geometry comprising an accretion disk component... along with a blackbody.'
  • domain assumption The X-ray dips are intrinsic dipping or absorption episodes of GX 13+1, and the dip versus non-dip comparison isolates a genuine change in emission geometry.
    The polarization-swing result depends on how dip intervals are selected and co-added; the abstract reports dips and swings without describing selection or background treatment.
  • standard math Standard IXPE polarization calibration and spectral fitting (Stokes I, Q, U with chi-squared model fitting) are correctly applied to the observations.
    Routine practice for IXPE analysis; cannot be verified from the abstract, and the body text is unavailable.

how reviews work

0 comments
Cite this review

Pith. "Pith review of X-ray and radio polarimetry of the neutron star low mass X-ray binary GX 13+1." pith.science (2026). https://pith.science/paper/GGVK7XFP

@misc{pith2026250805763,
  author       = {Pith},
  title        = {Pith review of: X-ray and radio polarimetry of the neutron star low mass X-ray binary GX 13+1},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GGVK7XFP}},
  note         = {Machine review of arXiv:2508.05763}
}
read the original abstract

We report the X-ray and radio polarization study of the neutron star (NS) low-mass X-ray binary (LMXB) GX 13+1 using the Imaging X-ray Polarimetry Explorer (IXPE) and Very Large Array (VLA). Simultaneous Neutron Star Interior Composition Explorer (NICER) observations show that the source was in parts of the Z state during our IXPE observations, exhibiting moderate changes in the hardness intensity diagram. The source exhibits X-ray dips in the light curve along with hints of polarization swings between the dip and non-dip states. The X-ray spectro-polarimetry results suggest a source geometry comprising an accretion disk component representing the softer disk emission, along with a blackbody representing the harder emission from the boundary layer (BL) or a spreading layer (SL). We investigate the geometry of GX 13+1 by considering our X-ray and radio polarization findings.

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

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

  1. What's the Buzz About GX 13+1? Constraining Coronal Geometry with QUEEN-BEE: A Bayesian Nested Sampling Framework for X-ray Polarization Rotation Analysis

    astro-ph.HE 2025-09 conditional novelty 6.0 of 10

    A new Bayesian framework recovers EVPA rotation in IXPE data and favors a slab coronal geometry for GX 13+1.

Reference graph

Works this paper leans on

33 extracted references · 4 canonical work pages · cited by 1 Pith paper

  1. [1]

    we formulate the BR-MPPI cost terms that are neces- sary to handle TT navigation applications and we show that the standard MPPI cannot solve similar problem arXiv:2508.05773v1 [cs.RO] 7 Aug 2025 with soft obstacle avoidance cost,

    arXiv 2025

  2. [2]

    we modify the BR-MPPI’s projection operator as a soft penalty, with penalty weights tuned using temporal- logic-robustness guided falsification,

    work page

  3. [3]

    we implement the framework in JAX enabling real- time execution on GPU platforms, and

    work page

  4. [4]

    for all {k}b k=a

    we evaluate the framework by simulating a pickup truck towing and parking a boat within a high-fidelity simulator (CarMaker). The first contribution is fundamental in solving the reverse maneuvers in tractor-trailer (TT) systems. As we elaborate in Section V, some constraints are captured in the cost function, while others, e.g., dynamic obstacles and hit...

    work page

  5. [5]

    Information-theoretic model predictive control: Theory and applica- tions to autonomous driving,

    G. Williams, P. Drews, B. Goldfain, J. M. Rehg, and E. A. Theodorou, “Information-theoretic model predictive control: Theory and applica- tions to autonomous driving,” IEEE Transactions on Robotics , vol. 34, no. 6, pp. 1603–1622, 2018

    2018

  6. [6]

    Control barrier function based quadratic programs for safety critical systems,

    A. D. Ames, X. Xu, J. W. Grizzle, and P. Tabuada, “Control barrier function based quadratic programs for safety critical systems,” IEEE Transactions on Automatic Control , vol. 62, no. 8, pp. 3861–3876, 2016

    2016

  7. [7]

    Model predictive contouring control,

    D. Lam, C. Manzie, and M. Good, “Model predictive contouring control,” in 49th IEEE Conference on Decision and Control (CDC) . IEEE, 2010, pp. 6137–6142

    2010

  8. [8]

    Model predictive control for autonomous driving: Comparing kinematic and dynamic models of tractor-trailer systems,

    G. Vallinder, J. Martensson, P. F. Lima, and S. Bhat, “Model predictive control for autonomous driving: Comparing kinematic and dynamic models of tractor-trailer systems,” in IEEE 27th International Confer- ence on Intelligent Transportation Systems (ITSC) , 2024, pp. 3154– 3159

    2024

Show all 33 references
  1. [9]

    Optimization-based au- tonomous racing of 1: 43 scale rc cars,

    A. Liniger, A. Domahidi, and M. Morari, “Optimization-based au- tonomous racing of 1: 43 scale rc cars,” Optimal Control Applications and Methods , vol. 36, no. 5, pp. 628–647, 2015

    2015

  2. [10]

    Neural configuration distance function for continuum robot control,

    K. Long, H. Parwana, G. Fainekos, B. Hoxha, H. Okamoto, and N. Atanasov, “Neural configuration distance function for continuum robot control,” arXiv:2409.13865, Tech. Rep., 2024

    2024 arXiv

  3. [11]

    Guaranteed-safe mppi through composite control barrier functions for efficient sampling in multi-constrained robotic systems,

    P. Rabiee and J. B. Hoagg, “Guaranteed-safe mppi through composite control barrier functions for efficient sampling in multi-constrained robotic systems,” arXiv:2410.02154, Tech. Rep., 2024

    2024 arXiv

  4. [12]

    Shield model predictive path integral: A computationally efficient robust mpc method using control barrier functions,

    J. Yin, C. Dawson, C. Fan, and P. Tsiotras, “Shield model predictive path integral: A computationally efficient robust mpc method using control barrier functions,” IEEE Robotics and Automation Letters , vol. 8, no. 11, pp. 7106–7113, 2023

    2023

  5. [13]

    Control barrier function augmentation in sampling-based control algorithm for sample efficiency,

    C. Tao, H. Kim, H. Yoon, N. Hovakimyan, and P. V oulgaris, “Control barrier function augmentation in sampling-based control algorithm for sample efficiency,” in American Control Conference (ACC) , 2022, pp. 3488–3493

    2022

  6. [14]

    MPPI-DBaS: Safe trajectory op- timization with adaptive exploration,

    F. Wang, Y . Cheng, and C. Tao, “MPPI-DBaS: Safe trajectory op- timization with adaptive exploration,” arXiv:2502.14387, Tech. Rep., 2025

    2025 arXiv

  7. [15]

    DualGuard MPPI: Safe and performant optimal control by combining sampling-based mpc and hamilton-jacobi reachability,

    J. Borquez, L. Raus, Y . U. Ciftci, and S. Bansal, “DualGuard MPPI: Safe and performant optimal control by combining sampling-based mpc and hamilton-jacobi reachability,” arXiv:2502.01924, Tech. Rep., 2025

    2025 arXiv

  8. [16]

    Adaptive robust path tracking preview control for tractor-trailer trucks considering trailer sway and stochastic disturbances,

    Y . Liu, M. Yue, X. Zhao, and G. Zong, “Adaptive robust path tracking preview control for tractor-trailer trucks considering trailer sway and stochastic disturbances,” IEEE Transactions on Intelligent Transportation Systems, vol. 26, no. 4, pp. 5422–5434, 2025

    2025

  9. [17]

    Motion planning using physics-informed lstms for autonomous driving,

    M. Selim, S. Bhat, and K. H. Johansson, “Motion planning using physics-informed lstms for autonomous driving,” in IEEE 27th In- ternational Conference on Intelligent Transportation Systems (ITSC) , 2024, pp. 2251–2258

    2024

  10. [18]

    Deep reinforcement learning based tractor-trailer tracking control,

    Q. Kang, A. Hartmannsgruber, S.-H. Tan, X. Zhang, and C.-M. Chew, “Deep reinforcement learning based tractor-trailer tracking control,” in 2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC) , 2024, pp. 3147–3153

    2024

  11. [19]

    The truck backer-upper: an example of self-learning in neural networks,

    D. Nguyen and B. Widrow, “The truck backer-upper: an example of self-learning in neural networks,” in International Joint Conference on Neural Networks , 1989, pp. 357–363 vol.2

    1989

  12. [20]

    Hybrid control of a truck and trailer vehicle,

    C. Altafini, A. Speranzon, and K. H. Johansson, “Hybrid control of a truck and trailer vehicle,” in Hybrid Systems: Computation and Control. Springer, 2002, pp. 21–34

    2002

  13. [21]

    Constrained backward path tracking control using a plug-in jackknife prevention system for autonomous tractor-trailers,

    M. Hejase, J. Jing, J. M. Maroli, Y . Bin Salamah, L. Fiorentini, and U. Ozguner, “Constrained backward path tracking control using a plug-in jackknife prevention system for autonomous tractor-trailers,” in 21st International Conference on Intelligent Transportation Systems (I...

    2018

  14. [22]

    Motion planning and model predictive control for automated tractor-trailer hitching maneuver,

    Z. Wang, A. Ahmad, R. Quirynen, Y . Wang, A. Bhagat, E. Zeino, Y . Zushi, and S. Di Cairano, “Motion planning and model predictive control for automated tractor-trailer hitching maneuver,” in IEEE Conference on Control Technology and Applications (CCTA) , 2022, pp. 676–682

    2022

  15. [23]

    Efficient safety-critical trajectory planning for any n-trailer system with ageneral model,

    L. Gao, B. Jia, D. Li, Y . Yang, and S. Xie, “Efficient safety-critical trajectory planning for any n-trailer system with ageneral model,” Control Engineering Practice , no. 158

  16. [24]

    Improved a- search guided tree for autonomous trailer planning,

    J. Leu, Y . Wang, M. Tomizuka, and S. D. Cairano, “Improved a- search guided tree for autonomous trailer planning,” in IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2022

    2022

  17. [25]

    Improved task and motion planning for rearrangement problems using optimal control*,

    A. Hellander, K. Bergman, and D. Axehill, “Improved task and motion planning for rearrangement problems using optimal control*,” in IEEE Intelligent V ehicles Symposium (IV) , 2024, pp. 2033–2040

    2024

  18. [26]

    Real- time optimization-based path planning for autonomous semi-trailer trucks*,

    P. Ma, L. Sun, S. Wei, D. Wan, F. Ding, D. Zhang, and S. Tian, “Real- time optimization-based path planning for autonomous semi-trailer trucks*,” in 2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC) , 2024, pp. 3133–3140

    2024

  19. [27]

    Apten- planner: Autonomous parking of semi-trailer train in extremely nar- row environments,

    M. Zhao, T. Shen, F. Wang, G. Yin, Z. Li, and Y . Zhang, “Apten- planner: Autonomous parking of semi-trailer train in extremely nar- row environments,” IEEE Transactions on Intelligent Transportation Systems, vol. 25, no. 5, pp. 4116–4132, 2024

    2024

  20. [28]

    Safety-critical control using optimal-decay control barrier function with guaranteed point- wise feasibility,

    J. Zeng, B. Zhang, Z. Li, and K. Sreenath, “Safety-critical control using optimal-decay control barrier function with guaranteed point- wise feasibility,” in 2021 American Control Conference (ACC). IEEE, 2021, pp. 3856–3863

    2021

  21. [29]

    Safe perception-based control under stochastic sensor uncertainty using conformal prediction,

    S. Yang, G. J. Pappas, R. Mangharam, and L. Lindemann, “Safe perception-based control under stochastic sensor uncertainty using conformal prediction,” in 62nd IEEE Conference on Decision and Control (CDC), 2023, pp. 6072–6078

    2023

  22. [30]

    Psy-taliro: A python toolbox for search-based test generation for cyber-physical systems,

    Q. Thibeault, J. Anderson, A. Chandratre, G. Pedrielli, and G. Fainekos, “Psy-taliro: A python toolbox for search-based test generation for cyber-physical systems,” 2021. [Online]. Available: https://arxiv.org/abs/2106.02200

    2021 arXiv

  23. [31]

    Mpcc++: Model predictive contouring control for time-optimal flight with safety con- straints,

    D. Kulic, G. Venture, K. Bekris, and E. Coronado, “Mpcc++: Model predictive contouring control for time-optimal flight with safety con- straints,” in Robotics: Science and Systems Conference (RSS 2024) , 2024

    2024

  24. [32]

    Hybrid a* for trailer truck,

    A. Sakai, “Hybrid a* for trailer truck,” https://github.com/yinflight/ HybridAStarTrailer, 2022, accessed: 2025-05-01

    2022

  25. [33]

    JAX: composable transformations of Python+NumPy programs,

    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. [Online]. Available: http://github. com/jax-ml/jax

    2018

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.