Pith. sign in

REVIEW 4 major objections 7 minor 47 references

Understanding Galaxy Morphology Evolution Through Cosmic Time via Redshift Conditioned Diffusion Models

T0 review · 4 major / 7 minor · reviewed 2026-08-12 · deepseek-v4-flash

Pith's one-line read A redshift-conditioned diffusion model learns galaxy morphology trends from images alone.

desk verdict A useful continuous-conditioning DDPM for galaxy simulation, but the physical-match claim rests on an undocumented metric pipeline and overclaimed novelty. read the letter →

arxiv 2411.18440 v2 pith:F4DLGK3Y submitted 2024-11-27 astro-ph.GA cs.CV

classification astro-ph.GAcs.CV
keywords galaxymorphologyredshiftdiffusionmodelsgenerativephotometricevolutionHyperSuprime-CamSérsicindex
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

This paper claims that a denoising diffusion model conditioned on a continuous redshift value can learn the joint structure linking a galaxy's distance to its appearance, and that the images it generates reproduce known morphology–redshift correlations. The authors train on Hyper Suprime-Cam galaxy images with spectroscopic redshifts and show that, across redshift bins, the mean ellipticity, semi-major axis, Sérsic index, and isophotal area of generated galaxies closely track the real test distribution. They also show that a CNN redshift predictor trained on real galaxies assigns redshifts to generated images that correlate with the conditioning redshift, indicating a genuine redshift signal in the output. If correct, this offers a path to photometric redshift estimation that uses morphological information, and a way to simulate physically plausible galaxy populations at arbitrary redshifts.

What carries the argument

The key mechanism is a redshift-conditioned denoising diffusion probabilistic model (DDPM) built on a U-Net, in which the conditioning variable is a continuous, log-transformed redshift z' = log(1+z) embedded via sinusoidal positional encoding along with the diffusion timestep. During training, Gaussian noise z' + N(0,σ) is added to the redshift value, forcing the model to interpolate smoothly across nearby redshifts rather than memorize discrete bins. This perturbation lets the model learn p(X|z) over a continuum, and the paper shows that a small perturbation (σ=0.1) yields physically accurate morphological metrics, whereas σ=0 (discrete conditioning) sharply degrades them.

What would settle it

Recompute all four morphological metrics on both real and generated images with a single, publicly documented pipeline using identical background subtraction, thresholding, PSF convolution, and deblending; if the redshift-binned means of the generated images then depart from the real means beyond the 95% confidence intervals shown in Figure 4, the paper's claim of physical fidelity is falsified. A second check: train a CNN redshift estimator on real images only, then verify that on generated images its predictions correlate with the conditioning redshift after matching the redshift distribution; a null result would directly contradict the claim that generated images carry a redshift signal.

Watch

Extended reading notes

Core claim

Continuous redshift conditioning enables a denoising diffusion model to learn the conditional distribution of galaxy images given distance, and the generated images reproduce the redshift-dependent trends in ellipticity, semi-major axis, Sérsic index, and isophotal area observed in real Hyper Suprime-Cam data. The authors train a U-Net DDPM on 64x64 five-band galaxy images with log-transformed redshift z' = log(1+z) as a continuous condition, adding Gaussian noise of standard deviation σ to the redshift during training. With σ=0.1 the model's generated images have morphological metric ratios of 0.93–0.98 relative to the real test-set means, outperforming discrete-conditioned baselines and continuous GANs on these physical metrics, despite a slightly worse FID than the discrete DDPM baseline. The paper interprets this as evidence that the model implicitly encodes physical galaxy structure and its evolution from photometry and redshift alone, without morphological labels.

Load-bearing premise

The morphological metrics computed on generated 64x64 five-band images are directly comparable to the same metrics on real HSC images, even though the paper does not describe the measurement software, background subtraction, thresholding, deblending, or PSF treatment used to derive ellipticity, semi-major axis, Sérsic index, and isophotal area.

Editorial extensions

If this is right

  • If correct, morphologically informed photometric redshift estimation becomes possible by inverting or conditioning such a generative model, potentially breaking degeneracies that pure color-based templates miss.
  • Astronomers could generate synthetic galaxy populations at any requested redshift for survey simulations, including redshifts where spectroscopic training data is scarce, without needing to tune physical parameters by hand.
  • The continuous conditioning scheme suggests that future galaxy generative models should treat physical parameters like redshift as continuous variables rather than discrete bins, since discretization degrades the physical fidelity of the output.
  • The benchmark table implies that FID is a poor standalone metric for scientific galaxy generation; evaluating physical morphology is necessary, and the paper offers a template for doing so.

Reading between the lines

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

  • A direct extension the paper does not pursue is to use the learned conditional model itself as a photometric-redshift estimator by searching for the conditioning redshift whose generated image matches a target image; the reported correlation between conditioning and CNN-predicted redshift suggests this inversion could be competitive, but the paper does not test it.
  • The smoothness of the learned conditional distribution could be quantified by generating a finely spaced redshift trajectory from a single seed and measuring how much the morphological metrics change between consecutive redshifts; the paper asserts smooth transitions but does not measure their continuity.
  • The comparison of real and generated morphological metrics would be more convincing if the measurement software were specified and applied identically to both sets; because that is absent, a follow-up using a fixed catalog-level pipeline would separate physical fidelity from pipeline consistency.
  • The model's increased scatter at high redshift, which the paper attributes to sparse training data, suggests a testable prediction: conditioning on ancillary physical variables (stellar mass, environment) alongside redshift should reduce that scatter if morphology evolution is driven by more than redshift alone.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 7 minor

Summary. The paper presents a denoising diffusion probabilistic model (DDPM) trained on 64x64 five-band Hyper Suprime-Cam galaxy cutouts, conditioned on a log-transformed continuous redshift, with Gaussian noise added to the conditioning value during training. The authors generate galaxies at test-set redshifts and compare them to real test galaxies using four morphological metrics: ellipticity, semi-major axis, Sérsic index, and isophotal area. They report close agreement of the metric distributions (Figure 3) and of mean trends versus redshift (Figure 4), and a strong correlation between a CNN-predicted redshift and the conditioning redshift (Figure 5). The paper additionally benchmarks against discrete- and continuous-conditioned GAN and DDPM baselines (Table 1) and presents visual comparisons. It claims to establish a 'first direct link between galaxy morphology and redshift' and to offer a path to redshift estimation from imaging data.

Significance. If the morphological agreement shown in Figures 3 and 4 holds under a properly specified measurement pipeline, the paper provides a useful demonstration that a redshift-conditioned generative model can learn morphology-redshift correlations from imaging alone, which is potentially valuable for simulation and survey validation. The continuous conditioning scheme with perturbation is a sensible technical contribution, and the ablation over the perturbation scale sigma (Table 1) is informative. However, the paper's broader claims (a 'first direct link' and a 'simple and effective path to redshift estimation') are not supported by the evidence, and the lack of a documented morphological measurement pipeline is a serious gap. The strengths are the public code release, the benchmark comparisons, and the explicit treatment of continuous conditioning.

major comments (4)
  1. [Section 4 (morphological metrics)] The central claim that the model reproduces morphology-z trends rests on the quantitative comparison in Figures 3 and 4, but the paper does not state how the four metrics are measured on either real or generated images. Ellipticity and isophotal area are threshold-dependent, the Sérsic index is strongly PSF-biased unless the PSF is modeled and convolved, and real HSC cutouts have sky noise, correlated pixel noise, and a wavelength-dependent PSF, whereas DDPM outputs are denoised from Gaussian noise and have different background statistics. Without specifying the detection threshold, background subtraction, deblending, PSF treatment, and band(s) used, the close match could be produced by the measurement pipeline rather than by the learned conditional distribution. Please provide a complete description of the metric pipeline (software and parameters), or adopt a public pipeline such as SExtractor/GALFIT/Statmorph, and validate the measurement on simulated images of known input parameters.
  2. [Abstract and Section 4 (Figure 5)] The redshift-prediction evidence in Figure 5 is partially circular and does not support the claim of a 'simple and effective path to redshift estimation'. Generated images are conditioned on the redshift z, so a predictor trained on real images will correlate with the conditioning value whenever the model has learned any z-dependent signal (e.g., overall color or flux), not necessarily morphology. The predictor CNN is from the authors' own prior work and may have been trained on the same data set, and the paper gives no comparison with photometric redshift baselines, no bias/scatter statistics, and no test on a completely independent sample. In addition, the abstract's phrase 'first study to establish a direct link between galaxy morphology and redshift' ignores the substantial existing literature on morphology-redshift relations (e.g., Conselice 2014); the correct claim is that a conditionally trained DDPM reproduces known trends.
  3. [Table 1] Table 1's physical benchmark columns are ratios of the average synthetic metric to the average real metric. A degenerate model that always outputs an image with the average ellipticity, size, Sérsic index, and area would score near 1 on all four columns while failing to match the joint distribution. The paper should supplement these global-mean ratios with per-redshift-bin distribution comparisons (e.g., Kolmogorov-Smirnov or energy distances) and report per-bin mean ratios with confidence intervals. This is necessary to support the claim of 'notably better performance in terms of the physical benchmarks' in Section 4.1.
  4. [Sections 2 and 5; Figures 3/4] Sections 2 and 5 state that 92.8% of the sample has z < 1.5 and that the model's performance is 'notably less reliable at higher redshifts', yet the title and abstract claim insight 'through cosmic time'. The redshift bins in Figure 4 should be shown with per-bin sample sizes and confidence intervals, and the high-z bins (z > 1.5) should either be demonstrated to reproduce the trends or explicitly excluded from the claim. Without this, the 'through cosmic time' claim is stronger than the evidence supports.
minor comments (7)
  1. [Abstract] The abstract contains grammatical issues: 'learns this missing joint structure, reproduces known morphology-z correlations' is missing 'and' before 'reproduces', and 'we verify on the HyperSuprime-Cam survey, that the model captures' contains a comma splice.
  2. [Section 1] Typos and incomplete phrasing appear: 'knowldedge', 'appoaches', and 'we explore the model's ability under to generate galaxies conditioned at different cosmic time-points' should be corrected.
  3. [Section 2] 'prohibitively compute and time-expensive' should be 'prohibitively computationally and temporally expensive'.
  4. [Section 4.1] There are several typos and grammatical errors: 'bechmark', 'HyperCam-Suprime', and 'Our model achieve the second best FID score our of the model that were compared' need revision.
  5. [Section 5] 'V AE's' and 'VQ-V AE's' should be 'VAEs' and 'VQ-VAEs', and 'Moreso' should be 'Moreover'.
  6. [References] The reference 'Lizarraga et al. 2024a' cited in the Introduction appears to be this same manuscript (arXiv:2411.18440); please replace it with the intended prior work or mark it as a self-citation appropriately.
  7. [Figure 6 caption] The cautionary paragraph about RGB projection and dust/heat artifacts is placed in the caption but would be clearer in the main text, and the sentence structures should be polished.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the morphology-z comparison rests on external HSC benchmarks and the redshift predictor is baseline-validated on real data, so no claim reduces to its own inputs by construction.

full rationale

The paper's central derivation is a conditional DDPM trained on external HSC image-redshift pairs; the morphological evaluation compares textbook metrics (ellipticity, semi-major axis, Sersic index, isophotal area) computed on real test images and on generated images, which is an external benchmark rather than a fitted or self-defined quantity. The redshift-prediction experiment uses a CNN predictor from the authors' prior work, but the paper explicitly verifies the predictor's baseline on real test galaxies (Fig. 5, left), so the predictor is independent support rather than a self-citation chain. Conditioning on z means that recovering z from generated images is a consistency check, but it is not statistically forced: a model that ignored the conditioning would fail it, and the predictor is fixed from real data. The paper itself flags limitations, including the dubious FID channel-subsampling comparison (Sec. 4.1) and physical caveats (Sec. 5), and self-citations for the train/test split and false-color convention are not load-bearing. No equation is defined in terms of its target, and no fitted parameter is renamed as a prediction. Therefore no circular step is exhibited.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The model is empirical and uses no fitted physical constants. The result rests on the DDPM equations (standard), on the assumption that photometric morphology carries redshift information (domain assumption), on the representativeness of the HSC spectroscopic sample, and on the ad hoc Gaussian perturbation of the conditioning redshift. No invented entities appear.

free parameters (3)
  • Redshift perturbation std dev sigma = 0.1
    Chosen by ablation over {0.1, 0.5, 1.0}; the best physical metric ratios in Table 1 occur at sigma=0.1, so the headline result depends on this hand-selected value.
  • Conditioning noise std dev = 0.01
    Added to the redshift embedding at each step; no sensitivity study reported.
  • Redshift log transform z' = log(1+z) = applied to all redshifts
    Introduced to keep the scale linear without validation; all evaluations use the transformed z.
assumptions (4)
  • standard math DDPM forward and reverse processes (Eqs. 1 and 2) model the data distribution
    Adopted from Ho et al. 2020 as background; not re-derived in this paper.
  • domain assumption Galaxy photometric morphology contains usable redshift information
    This is the paper's motivating claim; supported by prior literature, but the paper does not establish it independently.
  • domain assumption HSC spectroscopic sample represents galaxy populations at each redshift bin
    92.8% of galaxies are at z<1.5; high-z bins are sparse, acknowledged in Section 5.
  • ad hoc to paper Gaussian perturbation of the conditioning redshift yields smooth interpolation
    Central to the continuous-conditioning design; only empirical ablation is provided, with no formal guarantee or convergence test.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Understanding Galaxy Morphology Evolution Through Cosmic Time via Redshift Conditioned Diffusion Models." pith.science (2026). https://pith.science/paper/F4DLGK3Y

@misc{pith2026241118440,
  author       = {Pith},
  title        = {Pith review of: Understanding Galaxy Morphology Evolution Through Cosmic Time via Redshift Conditioned Diffusion Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/F4DLGK3Y}},
  note         = {Machine review of arXiv:2411.18440}
}
abstract

Redshift measures the distance to galaxies and underlies our understanding of the origin of the Universe and galaxy evolution. Spectroscopic redshift is the gold-standard method for measuring redshift, but it requires about $1000$ times more telescope time than broad-band imaging. That extra cost limits sky coverage and sample size and puts large spectroscopic surveys out of reach. Photometric redshift methods rely on imaging in multiple color filters and template fitting, yet they ignore the wealth of information carried by galaxy shape and structure. We demonstrate that a diffusion model conditioned on continuous redshift learns this missing joint structure, reproduces known morphology-$z$ correlations. We verify on the HyperSuprime-Cam survey, that the model captures redshift-dependent trends in ellipticity, semi-major axis, S\'ersic index, and isophotal area that these generated images correlate closely with true redshifts on test data. To our knowledge this is the first study to establish a direct link between galaxy morphology and redshift. Our approach offers a simple and effective path to redshift estimation from imaging data and will help unlock the full potential of upcoming wide-field surveys.

Figures

Figures reproduced from arXiv: 2411.18440 by the authors.

Figure 1
Figure 1. Source density versus r-band depth for major imaging and spectroscopic galaxy surveys. The r-band depth corresponds to the 5σ detection limit in a 2 ′′ aperture. Each point represents a survey, with completed imaging surveys (e.g., SDSS, DES, KiDS, HSC) shown in solid markers, and projected surveys (e.g., LSST, Euclid) shown in pink outlined squares. The y-axis shows the number of detected sources per square degree … view at source ↗
Figure 2
Figure 2. Model Architecture: Our model follows conventional DDPM implementations, but the [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. From left to right, the figure displays histograms comparing the frequency distribution of [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Mean morphological metrics as a function of redshift. Comparison between real test [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Redshift prediction quality for synthesized galaxies. (Left) CNN-predicted redshift [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Visual comparison of real and synthesized galaxies across redshifts. (Top row) Examples [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: DDPM (False Color) Galaxies Generated at non-cherry-picked redshift values. [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]
Figure 8
Figure 8. Figure 8: Real images and their corresponding denoised outputs for pertubed [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

47 extracted references · 27 canonical work pages

  1. [1]

    T. M. C. Abbott et al. Dark energy survey year 3 results: Cosmological constraints from galaxy clustering and weak lensing. Physical Review D, 105 0 (2): 0 023520, 2022

  2. [2]

    C. P. Ahn et al. The ninth data release of the sloan digital sky survey: First spectroscopic data from the sdss-iii baryon oscillation spectroscopic survey. The Astrophysical Journal Supplement Series, 203 0 (2): 0 21, 2012

  3. [3]

    Aihara et al

    H. Aihara et al. Second data release of the hyper suprime-cam subaru strategic program. Publications of the Astronomical Society of Japan, 71 0 (5): 0 114, 2019

  4. [4]

    Wasserstein gan

    Martin Arjovsky, Soumith Chintala, and L \'e on Bottou. Wasserstein gan. In International Conference on Machine Learning (ICML), pages 214--223, 2017

  5. [5]

    Conselice

    Christopher J. Conselice. The Evolution of Galaxy Structure Over Cosmic Time . Annual Review of Astronomy and Astrophysics, 52 0 (1): 0 291--337, 2014 a . doi:10.1146/annurev-astro-081913-040037

  6. [6]

    The evolution of galaxy structure over cosmic time

    Christopher J Conselice. The evolution of galaxy structure over cosmic time. Annual Review of Astronomy and Astrophysics, 52: 0 291--337, 2014 b

  7. [7]

    Diffusion Models Beat GANs on Image Synthesis

    Prafulla Dhariwal and Alexander Nichol. Diffusion Models Beat GANs on Image Synthesis . In M. Ranzato, A. Beygelzimer, Y. Dauphin, P.S. Liang, and J. Wortman Vaughan, editors, Advances in Neural Information Processing Systems, volume 34, pages 8780--8794. Curran Associates, Inc., 2021. URL https://proceedings.neurips.cc/paper_files/paper/2021/file/49ad23d...

  8. [8]

    Jane Wang

    Xin Ding, Yongwei Wang, Kao Zhang, and Z. Jane Wang. CCDM : Continuous conditional diffusion models for image generation, 2024

Show all 47 references
  1. [9]

    Galaxiesml: A dataset of galaxy images with matched redshifts

    Quynh Do, Andrew Lizarraga, Claudia Yeh, et al. Galaxiesml: A dataset of galaxy images with matched redshifts. Zenodo, 2024 a . URL https://zenodo.org/records/11117528

  2. [10]

    GalaxiesML: an imaging and photometric dataset of galaxies for machine learning , June 2024 b

    Tuan Do, Evan Jones, Bernie Boscoe, Yunqi (Billy) Li, and Kevin Alfaro. GalaxiesML: an imaging and photometric dataset of galaxies for machine learning , June 2024 b . URL https://doi.org/10.5281/zenodo.11117528

  3. [11]

    Taming transformers for high-resolution image synthesis

    Patrick Esser, Robin Rombach, and Bj \"o rn Ommer. Taming transformers for high-resolution image synthesis. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 12873--12883, 2021

  4. [12]

    Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio

    Ian J. Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial nets. In Z. Ghahramani, M. Welling, C. Cortes, N. Lawrence, and K.Q. Weinberger, editors, Advances in Neural Information P...

  5. [13]

    Gaussian error linear units (gelus)

    Dan Hendrycks and Kevin Gimpel. Gaussian error linear units (gelus). arXiv preprint arXiv:1606.08415, 2016

  6. [14]

    GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium

    Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, and Sepp Hochreiter. GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium . In Advances in Neural Information Processing Systems (NeurIPS), volume 30, pages 6626--6637, 2017

  7. [15]

    Second data release of the Hyper Suprime-Cam Subaru Strategic Program

    Makoto Ando Hiroaki Aihara, Yusra AlSayyad and et al. Second data release of the Hyper Suprime-Cam Subaru Strategic Program . Publications of the Astronomical Society of Japan, 71 0 (6): 0 114, 10 2019. ISSN 0004-6264. doi:10.1093/pasj/psz103. URL https://doi.org/10.1093/pasj/psz103

  8. [16]

    Denoising Diffusion Probabilistic Models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising Diffusion Probabilistic Models . In H. Larochelle, M. Ranzato, R. Hadsell, M.F. Balcan, and H. Lin, editors, Advances in Neural Information Processing Systems, volume 33, pages 6840--6851. Curran Associates, Inc., 2020 a . U...

  9. [17]

    Denoising diffusion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffusion probabilistic models. In Advances in Neural Information Processing Systems (NeurIPS), pages 6840--6851, 2020 b

  10. [18]

    Better prompt compression without multi-layer perceptrons, 2025

    Edouardo Honig, Andrew Lizarraga, Zijun Frank Zhang, and Ying Nian Wu. Better prompt compression without multi-layer perceptrons, 2025. URL https://arxiv.org/abs/2501.06730

  11. [19]

    Unlocking the potential of text-to-image diffusion with pac-bayesian theory, 2024

    Eric Hanchen Jiang, Yasi Zhang, Zhi Zhang, Yixin Wan, Andrew Lizarraga, Shufan Li, and Ying Nian Wu. Unlocking the potential of text-to-image diffusion with pac-bayesian theory, 2024. URL https://arxiv.org/abs/2411.17472

  12. [20]

    Analyzing and Improving the Training Dynamics of Diffusion Models

    Tero Karras, Miika Aittala, Jaakko Lehtinen, Janne Hellsten, Timo Aila, and Samuli Laine. Analyzing and Improving the Training Dynamics of Diffusion Models . In Proc. CVPR, 2024

  13. [21]

    Auto-encoding variational bayes

    Diederik P Kingma and Max Welling. Auto-encoding variational bayes. In International Conference on Learning Representations (ICLR), 2014

  14. [22]

    Latent plan transformer for trajectory abstraction: Planning as latent space inference

    Deqian Kong, Dehong Xu, Minglu Zhao, Bo Pang, Jianwen Xie, Andrew Lizarraga, Yuhao Huang, Sirui Xie, and Ying Nian Wu. Latent plan transformer for trajectory abstraction: Planning as latent space inference. In A. Globerson, L. Mackey, D. Belgrave, A. Fan, U. Paquet, J. Tomczak...

  15. [23]

    Kuijken et al

    K. Kuijken et al. The fourth data release of the kilo-degree survey. Astronomy & Astrophysics, 625: 0 A2, 2019

  16. [24]

    Deep generative models for galaxy image simulations

    François Lanusse, Rachel Mandelbaum, Siamak Ravanbakhsh, Chun-Liang Li, Peter Freeman, and Barnabás Póczos. Deep generative models for galaxy image simulations . Monthly Notices of the Royal Astronomical Society, 504 0 (4): 0 5543--5555, 05 2021. ISSN 0035-8711. doi:10.1093/mn...

  17. [25]

    Lastufka, M

    E. Lastufka, M. Drozdova, V. Kinakh, D. Piras, and S. Voloshynovskyy. Vision foundation models: can they be applied to astrophysics data?, 2024. URL https://arxiv.org/abs/2409.11175

  18. [26]

    Unredshift: A benchmark for learning galaxy morphology across cosmic time

    Minzhe Li, Andrew Lizarraga, et al. Unredshift: A benchmark for learning galaxy morphology across cosmic time. arXiv preprint arXiv:2401.12345, 2024 a

  19. [27]

    Using Galaxy Evolution as Source of Physics-Based Ground Truth for Generative Models , 2024 b

    Yun Qi Li, Tuan Do, Evan Jones, Bernie Boscoe, Kevin Alfaro, and Zooey Nguyen. Using Galaxy Evolution as Source of Physics-Based Ground Truth for Generative Models , 2024 b . URL https://arxiv.org/abs/2407.07229

  20. [28]

    Rl isn't enough

    Andrew Lizarraga. Rl isn't enough. https://drewrl3v.github.io/, April 2025. URL https://drewrl3v.github.io/blogs/april2025.html

  21. [29]

    Andrew Lizarraga, David Lee, Antoni Kubicki, Ashish Sahib, Elvis Nunez, Katherine Narr, and Shantanu H. Joshi. Alignment of tractography streamlines using deformation transfer via parallel transport. In Suheyla Cetin-Karayumak, Daan Christiaens, Matteo Figini, Pamela Guevara, ...

  22. [30]

    Narr, Kirsten A

    Andrew Lizarraga, Katherine L. Narr, Kirsten A. Donals, and Shantanu H. Joshi. Streamnet: A wae for white matter streamline analysis. In Erik Bekkers, Jelmer M. Wolterink, and Angelica Aviles-Rivero, editors, Proceedings of the First International Workshop on Geometric Deep Le...

  23. [31]

    Learning the evolution of physical structure of galaxies via diffusion models, 2024 a

    Andrew Lizarraga, Eric Hanchen Jiang, Jacob Nowack, Yun Qi Li, Ying Nian Wu, Bernie Boscoe, and Tuan Do. Learning the evolution of physical structure of galaxies via diffusion models, 2024 a . URL https://arxiv.org/abs/2411.18440

  24. [32]

    Andrew Lizarraga, Brandon Taraku, Edouardo Honig, Ying Nian Wu, and Shantanu H. Joshi. Differentiable vq-vae’s for robust white matter streamline encodings. In 2024 IEEE International Symposium on Biomedical Imaging (ISBI), pages 1--5, 2024 b . doi:10.1109/ISBI56570.2024.10635543

  25. [33]

    Decoupled weight decay regularization

    Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization. International Conference on Learning Representations (ICLR), 2020

  26. [34]

    Lotz, Patrik Jonsson, T

    Jennifer M. Lotz, Patrik Jonsson, T. J. Cox, and Joel R. Primack. Galaxy merger morphologies and time-scales from simulations of equal-mass gas-rich disc mergers. Monthly Notices of the Royal Astronomical Society, 391: 0 1137--1162, December 2008. ISSN 0035-8711. doi:10.1111/j...

  27. [35]

    Cosmic Star-Formation History

    Piero Madau and Mark Dickinson. Cosmic Star-Formation History . Annual Review of Astronomy and Astrophysics, 52: 0 415--486, August 2014. ISSN 0066-4146. doi:10.1146/annurev-astro-081811-125615

  28. [36]

    Detecting outliers in astronomical images with deep generative networks

    Berta Margalef-Bentabol, Marc Huertas-Company, Tom Charnock, Carla Margalef-Bentabol, Mariangela Bernardi, Yohan Dubois, Kate Storey-Fisher, and Lorenzo Zanisi. Detecting outliers in astronomical images with deep generative networks . Monthly Notices of the Royal Astronomical ...

  29. [37]

    Garcia, Jonah C

    Tri Nguyen, Francisco Villaescusa-Navarro, Siddharth Mishra-Sharma, Carolina Cuesta-Lazaro, Paul Torrey, Arya Farahi, Alex M. Garcia, Jonah C. Rose, Stephanie O'Neil, Mark Vogelsberger, Xuejian Shen, Cian Roche, Daniel Anglés-Alcázar, Nitya Kallivayalil, Julian B. Muñoz, Franc...

  30. [38]

    Improved denoising diffusion probabilistic models

    Alexander Quinn Nichol and Prafulla Dhariwal. Improved denoising diffusion probabilistic models. arXiv preprint arXiv:2102.09672, 2021

  31. [39]

    Nishizawa, Bau-Ching Hsieh, Masayuki Tanaka, and Tadafumi Takata

    Atsushi J. Nishizawa, Bau-Ching Hsieh, Masayuki Tanaka, and Tadafumi Takata. Photometric Redshifts for the Hyper Suprime-Cam Subaru Strategic Program Data Release 2 , 2020. URL https://arxiv.org/abs/2003.01511

  32. [40]

    Elvis Nunez, Andrew Lizarraga, and Shantanu H. Joshi. Srvfnet: A generative network for unsupervised multiple diffeomorphic functional alignment. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, pages 4481--4489, June 2021

  33. [41]

    Monitoring primitive interactions during the training of dnns

    Jie Ren, Xinhao Zheng, Jiyu Liu, Andrew Lizarraga, Ying Nian Wu, Liang Lin, and Quanshi Zhang. Monitoring primitive interactions during the training of dnns. Proceedings of the AAAI Conference on Artificial Intelligence, 39 0 (19): 0 20183--20191, Apr. 2025. doi:10.1609/aaai.v...

  34. [42]

    Realistic galaxy image simulation via score-based generative models

    Michael J Smith, James E Geach, Ryan A Jackson, Nikhil Arora, Connor Stone, and Stéphane Courteau. Realistic galaxy image simulation via score-based generative models . Monthly Notices of the Royal Astronomical Society, 511 0 (2): 0 1808--1818, 01 2022. ISSN 0035-8711. doi:10....

  35. [43]

    Deep unsupervised learning using nonequilibrium thermodynamics

    Jascha Sohl-Dickstein, Eric Weiss, Niru Maheswaranathan, and Surya Ganguli. Deep unsupervised learning using nonequilibrium thermodynamics. In Francis Bach and David Blei, editors, Proceedings of the 32nd International Conference on Machine Learning, volume 37 of Proceedings o...

  36. [44]

    Weiss, Niru Maheswaranathan, and Surya Ganguli

    Jascha Sohl-Dickstein, Eric A. Weiss, Niru Maheswaranathan, and Surya Ganguli. Deep unsupervised learning using nonequilibrium thermodynamics. In Proceedings of the 32nd International Conference on International Conference on Machine Learning - Volume 37, ICML'15, page 2256–22...

  37. [45]

    Denoising Diffusion Implicit Models

    Jiaming Song, Chenlin Meng, and Stefano Ermon. Denoising Diffusion Implicit Models . ArXiv, abs/2010.02502, 2020. URL https://api.semanticscholar.org/CorpusID:222140788

  38. [46]

    Patel, and Jeffrey Regier

    Zhiwei Xue, Yuhang Li, Yash J. Patel, and Jeffrey Regier. Diffusion Models for Probabilistic Deconvolution of Galaxy Images . ArXiv, abs/2307.11122, 2023. URL https://api.semanticscholar.org/CorpusID:260091385

  39. [47]

    Think twice before you act: Improving inverse problem solving with mcmc, 2024

    Yaxuan Zhu, Zehao Dou, Haoxin Zheng, Yasi Zhang, Ying Nian Wu, and Ruiqi Gao. Think twice before you act: Improving inverse problem solving with mcmc, 2024. URL https://arxiv.org/abs/2409.08551

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.