pith. sign in

arxiv: 2606.27021 · v1 · pith:IXS5BFQQnew · submitted 2026-06-25 · 🌌 astro-ph.IM

SMR: Scheduler with Multi-Channel Map-Encoded Reinforcement Learning for Radio Telescopes

Zhenyang Huang , Na Wang , Zhiyong Liu , Chuhao Gao This is my paper

Pith reviewed 2026-06-26 02:58 UTC · model grok-4.3

classification 🌌 astro-ph.IM
keywords radio telescope schedulingreinforcement learningAz-El mapsmulti-channel representationobservation schedulinginterference avoidancetime utilizationsingle-dish scheduler
0
0 comments X

The pith

Reinforcement learning with multi-channel Az-El maps improves radio telescope scheduling time utilization by about 10 percent over greedy baselines.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces SMR, a reinforcement learning approach to scheduling observations on large single-dish radio telescopes that must balance scientific return against mechanical and environmental limits. It converts lists of targets and constraints into multi-channel maps on an azimuth-elevation grid that carry both target properties and direction-specific signals such as interference risk and receiver gain. This map format supplies the learning agent with an explicit spatial view of the sky. Simulations drawn from real catalogs and site data show the method raises time utilization by roughly 10 percent relative to a tuned look-ahead greedy scheduler because the agent learns longer-horizon strategies. The three-channel version further improves interference and quality metrics over a standard neural-network baseline while preserving the utilization gains across 12-hour and 24-hour periods.

Core claim

SMR projects discrete targets onto an azimuth-elevation grid in the local horizon frame. The resulting aligned multi-channel sky maps encode target attributes together with direction-dependent cues such as satellite-interference risk and elevation-dependent receiver gain. This representation supplies an explicit spatial inductive bias that enables the reinforcement learning agent to learn directly from the sky state. Simulations based on real catalogs and site parameters show that SMR achieves about a 10 percent relative improvement in time utilization compared with a tuned look-ahead greedy baseline by learning non-myopic scheduling strategies. In the full three-channel setting SMR further

What carries the argument

The multi-channel Az-El map representation that projects targets onto a grid in the local horizon frame and encodes attributes together with direction-dependent cues such as interference risk and gain.

If this is right

  • The agent learns non-myopic strategies that raise overall time utilization.
  • The three-channel version enables simultaneous improvement in efficiency, interference avoidance, and observation quality.
  • The reported gains remain consistent across both 12-hour and 24-hour scheduling windows.
  • The method supplies a simple and extensible route to data-driven single-dish scheduling.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The map encoding could transfer to scheduling tasks in other fields that involve directional or spatial constraints.
  • Deployment on an actual telescope would reveal whether simulation gains survive real operational noise and changing conditions.
  • The same representation might combine with other reinforcement learning methods to push the interference and quality metrics higher.

Load-bearing premise

The multi-channel Az-El map representation supplies an explicit spatial inductive bias that lets the reinforcement learning agent learn the reported scheduling improvements directly from the sky state.

What would settle it

A controlled simulation that feeds the same reinforcement learning agent a flat list of targets instead of the multi-channel Az-El maps and measures whether the 10 percent utilization advantage over the greedy baseline disappears.

Figures

Figures reproduced from arXiv: 2606.27021 by Chuhao Gao, Na Wang, Zhenyang Huang, Zhiyong Liu.

Figure 1
Figure 1. Figure 1: Illustration of receiver gain curves as a function of elevation. In SMR, this information is embedded into Channel 3 to encourage observing targets at optimal elevations. Target i: α : 12h30m δ : +45◦ r : L-Band u : ROACH d : 10 min α, δ → Az, El Target Encoder (r, u, d) → R 1 Ch3: Gain gˆ(El)Map Ch2: Satellite-interference proxy Ch1: Target Azimuth (360◦) Elevation (10◦-80◦) Target i Map to (Azt, Elt) Fil… view at source ↗
Figure 2
Figure 2. Figure 2: Schematic of the Az–El map. Discrete target attributes are transformed and projected onto an azimuth–elevation grid to form a multi-channel spatial tensor that is ingested by the CNN-based policy network. 2.3.1. Final observation We use a CNN to encode Mt into a compact feature vector mt. The final observation fed to the policy is ot = [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: End-to-end framework of SMR. Discrete target attributes are projected to an Az–El horizon grid and encoded by a CNN, whose output is concatenated with the telescope state and passed to the SAC. All learnable components inside the box (CNN, feature fusion, and SAC) are optimized jointly end-to-end via gradient-based learning under the RL objective; the environment is used only to generate trajectories and r… view at source ↗
Figure 4
Figure 4. Figure 4: Utilization curves for SMR (solid) versus Greedy (dashed) across five time windows. Shaded regions indicate the standard deviation. SMR demonstrates greater stability and higher efficiency, particularly in 12-hour and 24-hour schedules [PITH_FULL_IMAGE:figures/full_fig_p010_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Reduction in cable unwrap events (Bars; Greedy - SMR) versus Efficiency Gain (Lines; Dual Axis) for 12h and 24h windows. Positive bars indicate that SMR triggered fewer cable unwraps than Greedy baseline. The scheduler effectively trades instrument switching costs for a reduction in time-consuming large-angle maneuvers [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
read the original abstract

Observation scheduling for large single-dish radio telescopes is a multi-objective optimization problem: schedulers must maximize on-source scientific return under strict mechanical and environmental constraints. Previous dynamic scheduling relies on expert-designed heuristics, while existing reinforcement-learning (RL) approaches often struggle with variable-length target lists and lack an intrinsic representation of sky geometry. We present SMR (Scheduler with Map-encoded Reinforcement Learning), which projects discrete targets onto an azimuth--elevation (Az--El) grid in the local horizon frame. The resulting aligned multi-channel sky maps encode target attributes together with direction-dependent cues such as satellite-interference risk and elevation-dependent receiver gain. This representation provides an explicit spatial inductive bias and enables SMR to learn directly from the sky state. Simulations based on real catalogs and site parameters show that, compared with a tuned look-ahead greedy baseline, SMR achieves about a 10\% relative improvement in time utilization by learning non-myopic scheduling strategies. In the full three-channel setting, SMR further achieves joint trade-off among efficiency, interference avoidance, and observation quality, with up to 17\% higher LIER and 54\% higher HGOR relative to an MLP baseline while maintaining higher utilization across both 12 h and 24 h horizons. Overall, SMR provides a simple and extensible way for data-driven single-dish scheduling.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript proposes SMR, a reinforcement learning scheduler for radio telescope observations that encodes the sky state using multi-channel azimuth-elevation maps incorporating target properties, interference risks, and gain factors. Through simulations with real catalogs, it claims a 10% improvement in time utilization over a look-ahead greedy baseline by learning non-myopic strategies, and in the three-channel configuration, superior performance in LIER (up to 17% higher) and HGOR (up to 54% higher) compared to an MLP baseline while preserving higher utilization over 12- and 24-hour periods.

Significance. If the simulation results are supported by rigorous methodology and statistical analysis, this work could provide a valuable extensible framework for multi-objective scheduling in single-dish radio astronomy, leveraging spatial representations to overcome limitations of traditional heuristics and standard RL approaches.

major comments (2)
  1. Abstract: The central performance claims (approximately 10% time utilization improvement, up to 17% higher LIER, up to 54% higher HGOR) are presented without any information on the RL training procedure, baseline tuning details, number of simulation runs, error bars, or statistical significance tests. This absence makes it impossible to determine whether the reported gains are robust or attributable to the proposed method.
  2. Abstract (method description): The assertion that the multi-channel Az-El map representation supplies an explicit spatial inductive bias enabling the agent to learn directly from the sky state and produce the stated gains is not supported by any ablation studies, component-wise comparisons, or analysis isolating the contribution of the map encoding versus other aspects of the RL formulation.
minor comments (1)
  1. Abstract: The acronyms LIER and HGOR appear without definition or expansion on first use.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments on result transparency and methodological support. We address each major comment point by point below, indicating planned revisions to strengthen the manuscript.

read point-by-point responses

  1. Referee: Abstract: The central performance claims (approximately 10% time utilization improvement, up to 17% higher LIER, up to 54% higher HGOR) are presented without any information on the RL training procedure, baseline tuning details, number of simulation runs, error bars, or statistical significance tests. This absence makes it impossible to determine whether the reported gains are robust or attributable to the proposed method.

    Authors: We agree that the abstract would benefit from greater transparency on evaluation methodology. The main text details the RL training procedure (Section 3), baseline tuning (Section 4.1), and presents results averaged over 50 independent simulation runs with standard deviations and paired t-tests for significance. To address the concern directly in the abstract, we will add the following concise clause: 'Metrics are reported as means over 50 runs with standard deviations; significance assessed via paired t-tests.' This revision ensures the claims are presented with appropriate context while respecting abstract length limits. revision: yes

  2. Referee: Abstract (method description): The assertion that the multi-channel Az-El map representation supplies an explicit spatial inductive bias enabling the agent to learn directly from the sky state and produce the stated gains is not supported by any ablation studies, component-wise comparisons, or analysis isolating the contribution of the map encoding versus other aspects of the RL formulation.

    Authors: The manuscript already includes a direct comparison of the full map-encoded SMR against an MLP baseline that lacks the spatial Az-El representation, showing the reported gains in LIER and HGOR. This comparison isolates the contribution of the map encoding to a degree. However, we acknowledge that dedicated ablation experiments (e.g., single-channel maps or non-spatial variants) would provide stronger isolation of the inductive bias. We will add a new subsection with these ablations, quantifying the performance impact of each map component on the utilization, LIER, and HGOR metrics. revision: yes

Circularity Check

0 steps flagged

No significant circularity identified

full rationale

The paper presents an RL-based scheduler using multi-channel Az-El map encoding and reports simulation-based performance gains (10% utilization improvement, higher LIER/HGOR) versus greedy and MLP baselines. No equations, fitted parameters, or derivation steps are shown that reduce any claimed prediction or result to its inputs by construction. The central claims rest on direct simulation outcomes from real catalogs rather than self-definitional mappings, renamed empirical patterns, or load-bearing self-citations. The spatial inductive bias is presented as a design choice whose effectiveness is evaluated externally via simulation, not assumed into the result.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The central claim rests on the unverified effectiveness of the map representation and RL training procedure; no explicit free parameters, axioms, or invented entities are listed in the abstract.

axioms (1)
  • domain assumption Projecting discrete targets onto an Az-El grid supplies useful spatial inductive bias for scheduling decisions
    Core premise of the described method
invented entities (1)
  • Multi-channel Az-El sky maps no independent evidence
    purpose: Encode target attributes together with direction-dependent cues for direct RL input
    New representation introduced to address variable-length target lists and sky geometry

pith-pipeline@v0.9.1-grok · 5777 in / 1153 out tokens · 61456 ms · 2026-06-26T02:58:51.890511+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

138 extracted references · 87 canonical work pages · 18 internal anchors

  1. [4]

    S., Bignell, C., Braatz, J., et al

    Balser, D. S., Bignell, C., Braatz, J., et al. 2009, title GBT Dynamic Scheduling System: Algorithms, Metrics, and Simulations, in Astronomical Society of the Pacific Conference Series, Vol. 411, Astronomical Data Analysis Software and Systems XVIII, ed. D. A. Bohlender, D. Durand, & P. Dowler (San Francisco: Astronomical Society of the Pacific), 330--333...

    2009

  2. [6]

    A., & Baan, W

    Fridman, P. A., & Baan, W. A. 2001, title RFI mitigation methods in radio astronomy, Astronomy & Astrophysics, 378, 327, 10.1051/0004-6361:20011166

    work page doi:10.1051/0004-6361:20011166 2001

  3. [8]

    2018, title Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor, in Proceedings of Machine Learning Research, Vol

    Haarnoja, T., Zhou, A., Abbeel, P., & Levine, S. 2018, title Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor, in Proceedings of Machine Learning Research, Vol. 80, Proceedings of the 35th International Conference on Machine Learning, ed. J. Dy & A. Krause (PMLR), 1861--1870. https://proceedings.mlr.press/v...

    2018

  4. [11]

    Hunter, J. D. 2007, title Matplotlib: A 2D graphics environment, Computing in Science & Engineering, 9, 90, 10.1109/MCSE.2007.55

    work page doi:10.1109/mcse.2007.55 2007

  5. [12]

    2007, title Dynamic Scheduling in ALMA , in Astronomical Society of the Pacific Conference Series, Vol

    Lucero, S., Farris, A., Kneissl, R., et al. 2007, title Dynamic Scheduling in ALMA , in Astronomical Society of the Pacific Conference Series, Vol. 376, Astronomical Data Analysis Software and Systems XVI, ed. R. A. Shaw, F. Hill, & D. J. Bell (San Francisco: Astronomical Society of the Pacific), 673--678. https://ui.adsabs.harvard.edu/abs/2007ASPC..376..673L

    2007

  6. [13]

    N., Hobbs, G

    Manchester , R. N., Hobbs , G. B., Teoh , A., & Hobbs , M. 2005, title The Australia Telescope National Facility Pulsar Catalogue, AJ, 129, 1993, 10.1086/428488

    work page internal anchor Pith review doi:10.1086/428488 2005

  7. [14]

    Human-level 21 control through deep reinforcement learning,

    Mnih, V., Kavukcuoglu, K., Silver, D., et al. 2015, title Human-level control through deep reinforcement learning, Nature, 518, 529. http://dx.doi.org/10.1038/nature14236

    work page doi:10.1038/nature14236 2015

  8. [15]

    E., Okumura, S

    Nyman, L.- ., Andreani, P., Hibbard, J. E., Okumura, S. K., & Wilson, R. W. 2010, title ALMA science operations, in Proc.\ SPIE, Vol. 7737, Observatory Operations: Strategies, Processes, and Systems III (Bellingham, WA: SPIE), 77370G, 10.1117/12.858023

    work page doi:10.1117/12.858023 2010

  9. [16]

    2000, A&AS, 143, 23, doi: 10.1051/aas:2000169

    Ochsenbein , F., Bauer , P., & Marcout , J. 2000, title The VizieR database of astronomical catalogues, A&AS, 143, 23, 10.1051/aas:2000169

    work page doi:10.1051/aas:2000169 2000

  10. [17]

    2009, title The GBT Dynamic Scheduling System: A New Scheduling Paradigm, in Astronomical Society of the Pacific Conference Series, Vol

    O'Neil, K., Balser, D., Bignell, C., et al. 2009, title The GBT Dynamic Scheduling System: A New Scheduling Paradigm, in Astronomical Society of the Pacific Conference Series, Vol. 411, Astronomical Data Analysis Software and Systems XVIII, ed. D. A. Bohlender, D. Durand, & P. Dowler (San Francisco: Astronomical Society of the Pacific), 147--154. https://...

    2009

  11. [18]

    2019, PyTorch: an imperative style, high-performance deep learning library (Red Hook, NY, USA: Curran Associates Inc.)

    Paszke, A., Gross, S., Massa, F., et al. 2019, PyTorch: an imperative style, high-performance deep learning library (Red Hook, NY, USA: Curran Associates Inc.). https://dl.acm.org/doi/10.5555/3454287.3455008

    work page doi:10.5555/3454287.3455008 2019

  12. [19]

    2011, title Scikit-learn: Machine Learning in P ython, Journal of Machine Learning Research, 12, 2825

    Pedregosa, F., Varoquaux, G., Gramfort, A., et al. 2011, title Scikit-learn: Machine Learning in P ython, Journal of Machine Learning Research, 12, 2825

    2011

  13. [20]

    2021, title Stable-Baselines3: Reliable Reinforcement Learning Implementations, Journal of Machine Learning Research, 22, 1

    Raffin, A., Hill, A., Gleave, A., et al. 2021, title Stable-Baselines3: Reliable Reinforcement Learning Implementations, Journal of Machine Learning Research, 22, 1. http://jmlr.org/papers/v22/20-1364.html

    2021

  14. [21]

    2025, sgp4 : Track Earth satellites given TLE data,, Python package, Python Package Index (PyPI) https://pypi.org/project/sgp4/

    Rhodes, B. 2025, sgp4 : Track Earth satellites given TLE data,, Python package, Python Package Index (PyPI) https://pypi.org/project/sgp4/

    2025

  15. [24]

    2023, Self-Driving Telescopes: Autonomous Scheduling of Astronomical Observation Campaigns with Offline Reinforcement Learning, 2311.18094

    Terranova, F., Voetberg, M., Nord, B., & Pagul, A. 2023, Self-Driving Telescopes: Autonomous Scheduling of Astronomical Observation Campaigns with Offline Reinforcement Learning, 2311.18094

    arXiv 2023

  16. [27]

    SciPy 1.0: Fundamental Algorithms for Scientific Computing in Python

    Virtanen, P., Gommers, R., Oliphant, T. E., et al. 2020, title SciPy 1.0: Fundamental Algorithms for Scientific Computing in Python , Nature Methods, 17, 261, 10.1038/s41592-019-0686-2

    work page doi:10.1038/s41592-019-0686-2 2020

  17. [29]

    2013, title The ultimate meteorological question from observational astronomers: how good is the cloud cover forecast? , 428, 3288, 10.1093/mnras/sts278

    Ye, Q.-Z., & Chen, S.-S. 2013, title The ultimate meteorological question from observational astronomers: how good is the cloud cover forecast? , 428, 3288, 10.1093/mnras/sts278

    work page doi:10.1093/mnras/sts278 2013

  18. [31]

    2018, title Introduction of measurement and control technology of FAST , in Ground-based and Airborne Telescopes VII, ed

    Zhu, L. 2018, title Introduction of measurement and control technology of FAST , in Ground-based and Airborne Telescopes VII, ed. H. K. Marshall & J. Spyromilio, Vol. 10700, International Society for Optics and Photonics (SPIE), 107001V, 10.1117/12.2309761

    work page doi:10.1117/12.2309761 2018

  19. [32]

    2025 , note =

    Rhodes, Brandon , title =. 2025 , note =

    2025

  20. [33]

    and Crawford, Paul and Hujsak, Richard and Kelso, T

    Vallado, David A. and Crawford, Paul and Hujsak, Richard and Kelso, T. S. , title = ". AIAA/AAS Astrodynamics Specialist Conference and Exhibit , year = 2006, month = aug, address =. doi:10.2514/6.2006-6753 , note =

    work page doi:10.2514/6.2006-6753 2006

  21. [34]

    doi:10.5281/zenodo.19343630 , url =

    HUANG, ZHENYANG , title =. doi:10.5281/zenodo.19343630 , url =

    work page doi:10.5281/zenodo.19343630

  22. [35]

    and Crawford, Paul , title = "

    Vallado, David A. and Crawford, Paul , title = ". AIAA/AAS Astrodynamics Specialist Conference and Exhibit , year = 2008, month = aug, address =. doi:10.2514/6.2008-6770 , note =

    work page doi:10.2514/6.2008-6770 2008

  23. [36]

    Ground-based and Airborne Telescopes VII , editor =

    Lichun Zhu , title =. Ground-based and Airborne Telescopes VII , editor =. 2018 , doi =

    2018

  24. [37]

    PyTorch: an imperative style, high-performance deep learning library , year =

    Paszke, Adam and Gross, Sam and Massa, Francisco and Lerer, Adam and Bradbury, James and Chanan, Gregory and Killeen, Trevor and Lin, Zeming and Gimelshein, Natalia and Antiga, Luca and Desmaison, Alban and K\". PyTorch: an imperative style, high-performance deep learning library , year =. Proceedings of the 33rd International Conference on Neural Informa...

  25. [38]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , year =

    Graham, Benjamin and Engelcke, Martin and van der Maaten, Laurens , title =. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , year =

  26. [39]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , year =

    Choy, Christopher and Gwak, JunYoung and Savarese, Silvio , title =. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , year =

  27. [40]

    Proceedings of the 35th International Conference on Machine Learning , pages =

    Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor , author =. Proceedings of the 35th International Conference on Machine Learning , pages =. 2018 , editor =

    2018

  28. [41]

    Journal of Machine Learning Research , year =

    Antonin Raffin and Ashley Hill and Adam Gleave and Anssi Kanervisto and Maximilian Ernestus and Noah Dormann , title =. Journal of Machine Learning Research , year =

  29. [42]

    and Farris, A

    Lucero, S. and Farris, A. and Kneissl, R. and others , title =. Astronomical Data Analysis Software and Systems XVI , editor =. 2007 , publisher =

    2007

  30. [43]

    and Balser, D

    O'Neil, K. and Balser, D. and Bignell, C. and Clark, M. and Condon, J. and McCarty, M. and Marganian, P. and Shelton, A. and Braatz, J. and Harnett, J. and Maddalena, R. and Mello, M. and Sessoms, E. , title =. Astronomical Data Analysis Software and Systems XVIII , editor =. 2009 , publisher =

    2009

  31. [44]

    Balser, D. S. and Bignell, C. and Braatz, J. and Clark, M. and Condon, J. and Harnett, J. and O'Neil, K. and Maddalena, R. and Marganian, P. and McCarty, M. and Sessoms, E. and Shelton, A. , title =. Astronomical Data Analysis Software and Systems XVIII , editor =. 2009 , publisher =

    2009

  32. [45]

    Experimental Astronomy , year =

    Van Der Schaaf, Kjeld and Broekema, Chris and Van Diepen, Ger and Van Meijeren, Ellen , title =. Experimental Astronomy , year =. doi:10.1007/s10686-005-2865-7 , url =

    work page doi:10.1007/s10686-005-2865-7

  33. [46]

    Astro2020 Science White Paper: The Next Decade of Astroinformatics and Astrostatistics

    The Next Decade of Astroinformatics and Astrostatistics. Bulletin of the American Astronomical Society , keywords =. doi:10.48550/arXiv.1903.06796 , archivePrefix =. 1903.06796 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1903.06796 1903

  34. [47]

    Nature , year = 2023, month = aug, volume =

    Scientific discovery in the age of artificial intelligence. Nature , year = 2023, month = aug, volume =. doi:10.1038/s41586-023-06221-2 , adsurl =

    work page doi:10.1038/s41586-023-06221-2 2023

  35. [48]

    Science China Physics, Mechanics, and Astronomy , keywords =

    The Qitai radio telescope. Science China Physics, Mechanics, and Astronomy , keywords =. doi:10.1007/s11433-023-2131-1 , archivePrefix =. 2310.07163 , primaryClass =

    work page doi:10.1007/s11433-023-2131-1

  36. [49]

    Approximation capabilities of multilayer feedforward networks , journal =

    Kurt Hornik , keywords =. Approximation capabilities of multilayer feedforward networks , journal =. 1991 , issn =. doi:https://doi.org/10.1016/0893-6080(91)90009-T , url =

    work page doi:10.1016/0893-6080(91)90009-t 1991

  37. [50]

    2024 , eprint=

    Convolutional Kolmogorov-Arnold Networks , author=. 2024 , eprint=

    2024

  38. [51]

    2024 , eprint=

    KAN: Kolmogorov-Arnold Networks , author=. 2024 , eprint=

    2024

  39. [52]

    Deep Residual Learning for Image Recognition , year=

    He, Kaiming and Zhang, Xiangyu and Ren, Shaoqing and Sun, Jian , booktitle=. Deep Residual Learning for Image Recognition , year=

  40. [53]

    Attention is All you Need , url =

    Vaswani, Ashish and Shazeer, Noam and Parmar, Niki and Uszkoreit, Jakob and Jones, Llion and Gomez, Aidan N and Kaiser, ukasz and Polosukhin, Illia , booktitle =. Attention is All you Need , url =

  41. [54]

    2023 , month =

    He Jia and Hong-Ming Zhu and Ue-Li Pen , title =. 2023 , month =. doi:10.3847/1538-4357/aca8aa , url =

    work page doi:10.3847/1538-4357/aca8aa 2023

  42. [55]

    2022 , month =

    Yan-Ling Liu and Jian Li and Zhi-Yong Liu and Mao-Zheng Chen and Jian-Ping Yuan and Na Wang and Rai Yuen and Hao Yan , title =. 2022 , month =. doi:10.1088/1674-4527/ac833a , url =

    work page doi:10.1088/1674-4527/ac833a 2022

  43. [56]

    2024 , eprint=

    Self-Supervised Learning on MeerKAT Wide-Field Continuum Images , author=. 2024 , eprint=

    2024

  44. [57]

    2022 , eprint=

    Strong Gravitational Lensing Parameter Estimation with Vision Transformer , author=. 2022 , eprint=

    2022

  45. [59]

    and Bottou, L

    Lecun, Y. and Bottou, L. and Bengio, Y. and Haffner, P. , journal=. Gradient-based learning applied to document recognition , year=

  46. [60]

    ImageNet Classification with Deep Convolutional Neural Networks , url =

    Krizhevsky, Alex and Sutskever, Ilya and Hinton, Geoffrey E , booktitle =. ImageNet Classification with Deep Convolutional Neural Networks , url =

  47. [61]

    2022 , eprint=

    Galaxy Morphological Classification with Efficient Vision Transformer , author=. 2022 , eprint=

    2022

  48. [62]

    Unsupervised Galaxy Morphological Visual Representation with Deep Contrastive Learning , volume=

    Wei, Shoulin and Li, Yadi and Lu, Wei and Li, Nan and Liang, Bo and Dai, Wei and Zhang, Zhijian , year=. Unsupervised Galaxy Morphological Visual Representation with Deep Contrastive Learning , volume=. Publications of the Astronomical Society of the Pacific , publisher=. doi:10.1088/1538-3873/aca04e , number=

    work page doi:10.1088/1538-3873/aca04e

  49. [63]

    The structure and assembly history of cluster-size haloes in Self-Interacting Dark Matter

    Brinckmann, Thejs and Zavala, Jes\'us and Rapetti, David and Hansen, Steen H. and Vogelsberger, Mark. The structure and assembly history of cluster-sized haloes in self-interacting dark matter. Mon. Not. Roy. Astron. Soc. 2018. doi:10.1093/mnras/stx2782. arXiv:1705.00623

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/stx2782 2018

  50. [64]

    and Schaye, Joop and Bird, Simeon and Le Brun, Amandine M

    McCarthy, Ian G. and Schaye, Joop and Bird, Simeon and Le Brun, Amandine M. C. , title = ". Monthly Notices of the Royal Astronomical Society , volume =. 2016 , month =. doi:10.1093/mnras/stw2792 , url =

    work page doi:10.1093/mnras/stw2792 2016

  51. [65]

    C. L. Bennett and D. Larson and J. L. Weiland and N. Jarosik and G. Hinshaw and N. Odegard and K. M. Smith and R. S. Hill and B. Gold and M. Halpern and E. Komatsu and M. R. Nolta and L. Page and D. N. Spergel and E. Wollack and J. Dunkley and A. Kogut and M. Limon and S. S. Meyer and G. S. Tucker and E. L. Wright , title =. 2013 , month =. doi:10.1088/00...

    work page doi:10.1088/0067-0049/208/2/20 2013

  52. [66]

    , keywords =

    Springel, Volker , title = ". Monthly Notices of the Royal Astronomical Society , volume =. 2005 , month =. doi:10.1111/j.1365-2966.2005.09655.x , url =

    work page doi:10.1111/j.1365-2966.2005.09655.x 2005

  53. [67]

    Schaye, Joop and Vecchia, Claudio Dalla and Booth, C. M. and Wiersma, Robert P. C. and Theuns, Tom and Haas, Marcel R. and Bertone, Serena and Duffy, Alan R. and McCarthy, I. G. and van de Voort, Freeke , title = ". Monthly Notices of the Royal Astronomical Society , volume =. 2010 , month =. doi:10.1111/j.1365-2966.2009.16029.x , url =

    work page doi:10.1111/j.1365-2966.2009.16029.x 2010

  54. [68]

    Wiersma, Robert P. C. and Schaye, Joop and Smith, Britton D. , title = ". Monthly Notices of the Royal Astronomical Society , volume =. 2009 , month =. doi:10.1111/j.1365-2966.2008.14191.x , url =

    work page doi:10.1111/j.1365-2966.2008.14191.x 2009

  55. [69]

    , keywords =

    Dalla Vecchia, Claudio and Schaye, Joop. Simulating galactic outflows with kinetic supernova feedback. Mon. Not. Roy. Astron. Soc. 2008. doi:10.1111/j.1365-2966.2008.13322.x. arXiv:0801.2770

    work page doi:10.1111/j.1365-2966.2008.13322.x 2008

  56. [70]

    2011 , eprint=

    Euclid Definition Study Report , author=. 2011 , eprint=

    2011

  57. [71]

    and Massey, Richard

    Harvey, David R. and Massey, Richard. Weak gravitational lensing measurements of Abell\,2744 using JWST and shear measurement algorithm pyRRG-JWST. Mon. Not. Roy. Astron. Soc. 2024. doi:10.1093/mnras/stae370. arXiv:2401.16478

    work page doi:10.1093/mnras/stae370 2024

  58. [72]

    2016 , eprint=

    The design and development of a high-resolution visible-to-near-UV telescope for balloon-borne astronomy: SuperBIT , author=. 2016 , eprint=

    2016

  59. [73]

    , keywords =

    Schaye, Joop and Dalla Vecchia, Claudio. On the relation between the Schmidt and Kennicutt-Schmidt star formation laws and its implications for numerical simulations. Mon. Not. Roy. Astron. Soc. 2008. doi:10.1111/j.1365-2966.2007.12639.x. arXiv:0709.0292

    work page doi:10.1111/j.1365-2966.2007.12639.x 2008

  60. [74]

    Wiersma, Robert P. C. and Schaye, Joop and Theuns, Tom and Dalla Vecchia, Claudio and Tornatore, Luca. Chemical enrichment in cosmological, smoothed particle hydrodynamics simulations. Mon. Not. Roy. Astron. Soc. 2009. doi:10.1111/j.1365-2966.2009.15331.x. arXiv:0902.1535

    work page doi:10.1111/j.1365-2966.2009.15331.x 2009

  61. [75]

    The BAHAMAS project: the CMB--large-scale structure tension and the roles of massive neutrinos and galaxy formation

    Mccarthy, Ian G. and Bird, Simeon and Schaye, Joop and Harnois-Deraps, Joachim and Font, Andreea S. and Van Waerbeke, Ludovic. The BAHAMAS project: the CMB large-scale structure tension and the roles of massive neutrinos and galaxy formation. Mon. Not. Roy. Astron. Soc. 2018. doi:10.1093/mnras/sty377. arXiv:1712.02411

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/sty377 2018

  62. [76]

    and Dambre, Joni , title = "

    Dieleman, Sander and Willett, Kyle W. and Dambre, Joni , title = ". Monthly Notices of the Royal Astronomical Society , volume =. 2015 , month =. doi:10.1093/mnras/stv632 , url =

    work page doi:10.1093/mnras/stv632 2015

  63. [78]

    arXiv preprint arXiv:1707.06347 , year=

    Proximal policy optimization algorithms , author=. arXiv preprint arXiv:1707.06347 , year=

    Pith/arXiv arXiv

  64. [79]

    International conference on machine learning , pages=

    Trust region policy optimization , author=. International conference on machine learning , pages=. 2015 , organization=

    2015

  65. [80]

    Caminha and R

    Massimo Meneghetti and Guido Davoli and Pietro Bergamini and Piero Rosati and Priyamvada Natarajan and Carlo Giocoli and Gabriel B. Caminha and R. Benton Metcalf and Elena Rasia and Stefano Borgani and Francesco Calura and Claudio Grillo and Amata Mercurio and Eros Vanzella , title =. Science , volume =. 2020 , doi =. https://www.science.org/doi/pdf/10.11...

    work page doi:10.1126/science.aax5164 2020

  66. [81]

    Monthly Notices of the Royal Astronomical Society , volume =

    Gilman, Daniel and Birrer, Simon and Nierenberg, Anna and Treu, Tommaso and Du, Xiaolong and Benson, Andrew , title = ". Monthly Notices of the Royal Astronomical Society , volume =. 2019 , month =. doi:10.1093/mnras/stz3480 , url =

    work page doi:10.1093/mnras/stz3480 2019

  67. [82]

    Warren and Kevork Abazajian and Daniel E

    Michael S. Warren and Kevork Abazajian and Daniel E. Holz and Luís Teodoro , title =. The Astrophysical Journal , abstract =. 2006 , month =. doi:10.1086/504962 , url =

    work page doi:10.1086/504962 2006

  68. [83]

    The Astrophysical Journal , abstract =

    Danilo Marchesini and Elena D’Onghia and Guido Chincarini and Claudio Firmani and Paolo Conconi and Emilio Molinari and Andrea Zacchei , title =. The Astrophysical Journal , abstract =. 2002 , month =. doi:10.1086/341475 , url =

    work page doi:10.1086/341475 2002

  69. [84]

    and Navarro, Julio F

    Oman, Kyle A. and Navarro, Julio F. and Fattahi, Azadeh and Frenk, Carlos S. and Sawala, Till and White, Simon D. M. and Bower, Richard and Crain, Robert A. and Furlong, Michelle and Schaller, Matthieu and Schaye, Joop and Theuns, Tom , title = ". Monthly Notices of the Royal Astronomical Society , volume =. 2015 , month =. doi:10.1093/mnras/stv1504 , url =

    work page doi:10.1093/mnras/stv1504 2015

  70. [85]

    Observational Evidence for Self-Interacting Cold Dark Matter , author =. Phys. Rev. Lett. , volume =. 2000 , month =. doi:10.1103/PhysRevLett.84.3760 , url =

    work page doi:10.1103/physrevlett.84.3760 2000

  71. [86]

    , keywords =

    The Aquarius Project: the subhaloes of galactic haloes. Monthly Notices of the Royal Astronomical Society , keywords =. doi:10.1111/j.1365-2966.2008.14066.x , archivePrefix =. 0809.0898 , primaryClass =

    work page doi:10.1111/j.1365-2966.2008.14066.x 2008

  72. [87]

    Y., V¨ ais¨ anen, P., Whitelock, P

    Resolving cosmic structure formation with the Millennium-II Simulation. Monthly Notices of the Royal Astronomical Society , keywords =. doi:10.1111/j.1365-2966.2009.15191.x , archivePrefix =. 0903.3041 , primaryClass =

    work page doi:10.1111/j.1365-2966.2009.15191.x 2009

  73. [88]

    H., Bullock J

    Weinberg, David H. and Bullock, James S. and Governato, Fabio and Kuzio de Naray, Rachel and Peter, Annika H. G. , year=. Cold dark matter: Controversies on small scales , volume=. Proceedings of the National Academy of Sciences , publisher=. doi:10.1073/pnas.1308716112 , number=

    work page doi:10.1073/pnas.1308716112

  74. [89]

    Planck Collaboration et al.A&A, 641:A6, Sept

    Firmani, C. and D'Onghia, E. and Avila-Reese, V. and Chincarini, G. and Hernández, X. , title = ". Monthly Notices of the Royal Astronomical Society , volume =. 2000 , month =. doi:10.1046/j.1365-8711.2000.03555.x , url =

    work page doi:10.1046/j.1365-8711.2000.03555.x 2000

  75. [90]

    Observational and Theoretical Constraints on Singular Dark Matter Halos

    Observational and Theoretical Constraints on Singular Dark Matter Halos. , keywords =. doi:10.1086/187350 , archivePrefix =. astro-ph/9402004 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1086/187350

  76. [91]

    , keywords =

    The Galaxy Galaxy Strong Lensing Cross Section and the Internal Distribution of Matter in CDM Substructure. , keywords =. doi:10.3847/1538-4357/ad51fd , archivePrefix =. 2404.16951 , primaryClass =

    work page doi:10.3847/1538-4357/ad51fd

  77. [92]

    , keywords =

    A persistent excess of galaxy-galaxy strong lensing observed in galaxy clusters. , keywords =. doi:10.1051/0004-6361/202346975 , archivePrefix =. 2309.05799 , primaryClass =

    work page doi:10.1051/0004-6361/202346975

  78. [93]

    A critical assessment of solutions to the galaxy diversity problem

    Zentner, Aidan and Dandavate, Siddharth and Slone, Oren and Lisanti, Mariangela. A critical assessment of solutions to the galaxy diversity problem. JCAP. 2022. doi:10.1088/1475-7516/2022/07/031. arXiv:2202.00012

    work page doi:10.1088/1475-7516/2022/07/031 2022

  79. [94]

    Monthly Notices of the Royal Astronomical Society , volume =

    Santos-Santos, Isabel M E and Navarro, Julio F and Robertson, Andrew and Benítez-Llambay, Alejandro and Oman, Kyle A and Lovell, Mark R and Frenk, Carlos S and Ludlow, Aaron D and Fattahi, Azadeh and Ritz, Adam , title = ". Monthly Notices of the Royal Astronomical Society , volume =. 2020 , month =. doi:10.1093/mnras/staa1072 , url =

    work page doi:10.1093/mnras/staa1072 2020

  80. [95]

    Self-interacting inelastic dark matter: A viable solution to the small scale structure problems

    Self-interacting inelastic dark matter: a viable solution to the small scale structure problems. , keywords =. doi:10.1088/1475-7516/2017/03/048 , archivePrefix =. 1612.06681 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1088/1475-7516/2017/03/048 2017

Showing first 80 references.