Pith. sign in

REVIEW 3 major objections 6 minor 273 references

AGNFormer I: Reconstruction of AGN spectra using a probabilistic transformer model

T0 review · 3 major / 6 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read A transformer trained on SDSS quasar spectra can reconstruct masked broad emission lines to within 4–16% of the flux, and outperforms purpose-built algorithms on a blind reconstruction challenge.

desk verdict Solid self-supervised spectral reconstruction with a clean transformer design, but the X-Shooter win may be object leakage rather than cross-resolution transfer. read the letter →

arxiv 2607.19640 v1 pith:RP5PFWCU submitted 2026-07-22 astro-ph.GA astro-ph.IM

Benedict L. Rouse , Franz E. Bauer , Tomasz Różański , Guillermo Cabrera-Vives , Patricia Arévalo , Cristóbal Rodrigo Donoso Oliva , Pavlos Protopapas , Gordon T. Richards
show 3 more authors
Paula Sánchez-Sáez Ezequiel Treister Christian Wolf
This is my paper
classification astro-ph.GAastro-ph.IM
keywords activegalacticnucleiquasarspectratransformermaskedautoencoderspectralreconstructionbroad-lineregionuncertaintycalibrationself-supervisedlearning
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 argues that a single transformer trained with a masked-autoencoding objective on the full SDSS DR16Q quasar catalogue can reconstruct masked broad emission lines (Hα, Hβ, MgII, CIV, Lyα, and others) using only the surrounding continuum and weak features. On held-out spectra the reconstruction errors are 10–16% of the flux for the full test set and 4–8% for high signal-to-noise (S/N>10) spectra, with an intrinsic floor of roughly 2–6% at S/N≈40; even larger masked halves are recovered to 12–25% (5–15%). The same untuned model wins a blind challenge for the Lyα region against eleven purpose-built reconstruction algorithms on both SDSS and X-Shooter data, demonstrating resolution-invariant transfer. If correct, this shows that the continuum and weak lines encode enough information to determine the broad-line region, enabling AGN property extraction (e.g., black hole masses) even when the main lines are corrupted, absorbed, or outside the observed window.

What carries the argument

AGNFormer is a transformer encoder trained with a masked-autoencoding objective. Three design choices carry the argument: (1) a wavelength-based positional encoder substituting log10λ for token indices, making the model insensitive to pixel grid and spectral resolution; (2) an uncertainty-aware head predicting both mean flux and variance, with a heteroscedastic Gaussian negative-log-likelihood loss weighting each pixel by its observational error (input fluxes are noise-perturbed during training); (3) a masking scheme combining random pixel, random chunk, and fixed broad-line masks (35:40:25), plus a Euclidean normalization computed only over unmasked pixels to prevent leakage. Together these

What would settle it

Evaluate the fully trained model on a set of ~100 quasar spectra from a different instrument covering the same rest-frame lines, without interpolating to the SDSS pixel grid, and compute the predicted/observed flux ratios over the full ±10,000 km/s windows; if the median error exceeds the quoted 4–8% for S/N>10, the generalization claim is falsified. Alternatively, retrain with noise values shuffled randomly across pixels: a sharp drop in reconstruction accuracy would indicate the model uses noise-pattern information rather than spectral physics.

Watch

Extended reading notes

Core claim

AGNFormer is an encoder-only transformer trained self-supervised on ~750,000 SDSS DR16Q quasar spectra to predict both mean flux and per-pixel variance of masked regions. The central claim is that the broad-line region is learnable from the rest of the spectrum: with major broad lines masked to ±10,000 km/s, it reproduces unseen spectra to 10–16% (full) and 4–8% (S/N>10) of the flux, with well-calibrated uncertainties. On a blind Lyα reconstruction challenge it beats all eleven purpose-built algorithms (MAE 0.017 X-Shooter, 0.022 SDSS) despite being trained only on SDSS at R≈2000 and tested on higher-resolution X-Shooter data without fine-tuning. The authors conclude that continuum and weak

Load-bearing premise

The headline accuracy is measured on a random split of the same SDSS DR16Q catalogue used for training, so training and test spectra share the survey's noise, calibration, and redshift-assignment systematics; if the model exploits those systematics rather than intrinsic AGN correlations, the reported percentages overstate performance on genuinely new data.

Editorial extensions

If this is right

  • If the broad-line region is this well constrained by the continuum and weak features, black hole masses and Eddington ratios—normally measured from broad-line widths—can be estimated from spectra where the lines are corrupted, absorbed, or outside the observed window.
  • The wavelength-based positional encoder offers a concrete architectural path to cross-instrument, cross-resolution transfer: a model trained at R≈2000 generalizes to R≈4500–7500 without retuning, as shown on the X-Shooter part of the blind challenge.
  • The self-supervised pretraining on ~750,000 spectra yields representations that the authors argue will support downstream fine-tuning tasks, including probabilistic prediction of AGN properties and variability studies.
  • For broad-absorption-line quasars, the model—trained without BALs—reconstructs an intrinsic CIV profile and overpredicts flux blueward of the line, a behavior the authors propose as a tool for recovering intrinsic line properties in absorbed sources.
  • Whole-half predictions extend the effective wavelength coverage of a spectrograph, offering a way to homogenize heterogeneous surveys and to flag strong features in unobserved windows for follow-up.

Reading between the lines

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

  • A clear testable extension is to train on a second, independent spectroscopic survey with different instrumentation and noise properties, and measure whether the reported in-survey accuracy survives; the current external evidence rests on only 30 X-Shooter spectra compared over a 52 Å window.
  • The error floor of ~2–6% at S/N≈40 suggests the residual error is dominated by physical diversity (e.g., NLR geometry, orientation, absorption) rather than noise; improvements may require multi-epoch or multi-band conditioning, not just more spectra.
  • The per-pixel variance output could itself serve as an anomaly detector: large predicted uncertainty on a cleanly measured flux would flag genuinely ambiguous regions (e.g., absorption troughs or unusual profiles) for follow-up.
  • If the continuum-to-BLR mapping is as tight as claimed, reverberation-mapping campaigns could use the model's prediction as a prior or interpolant between sparse epochs, potentially reducing the number of epochs needed to constrain BLR geometry.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. The paper introduces AGNFormer, an encoder-only transformer trained with a masked-autoencoding objective on SDSS DR16Q rest-frame quasar spectra. The inputs are wavelength, flux, and reported flux uncertainty; two projection heads output a predictive Gaussian mean and variance for each masked pixel, trained with a heteroscedastic negative log-likelihood. Fixed ±10^4 km/s masks around major broad lines, plus half-spectrum masks, are evaluated on an unseen 3% test split. The authors report 10–16% (4–8%) median predicted/observed dispersion for full (S/N>10) broad-line reconstruction, an error floor of ≈2–6% at S/N≈40, calibrated pull distributions, and QMS-resolved reconstructions including FeII complexes and Civ profiles. They also benchmark against the Greig et al. (2024) Lyα reconstruction challenge, reporting superior MAE on 30 X-Shooter and 30 SDSS spectra.

Significance. If the in-survey results hold, the paper provides a useful, well-calibrated self-supervised spectral reconstruction method. Its strengths include a held-out test split, S/N-resolved pull diagnostics, an explicit error-floor fit, an external benchmark comparison, public code and pretrained weights, a formal positional-encoder equivalence proof in Appendix A, and a clear statement that the predictive distribution is an independent Gaussian. However, the strongest external generalization claim rests on the X-Shooter comparison, which is compromised by a likely object-level overlap with the training set, and the quantitative half-spectrum claim in the abstract is not supported by any measurement in Section 5.2. The in-survey broad-line reconstruction numbers are credible, but the cross-resolution and half-spectrum headline claims need substantial revision.

major comments (3)
  1. [§6.1, Table 2] The X-Shooter benchmark likely leaks object identity. The model is 'pretrained on the entire SDSS DR16Q catalogue' and tested on 30 X-Shooter spectra from XQ-100. The text excludes only 'the same 30 SDSS (not present in our training sample)' and does not state that the SDSS spectra of the X-Shooter targets were removed from training. XQ-100 is an SDSS-selected sample at z≈3.5–4.5, so those objects almost certainly have their own SDSS spectra inside DR16Q; if so, the model has seen each test object's continuum and line correlations at SDSS resolution. The Table 2 value MAE=0.017 and the 'resolution invariance' conclusion would then reflect object memorization, not cross-resolution generalization to unseen objects. Please verify object overlap explicitly, rerun with overlapping SDSS spectra excluded, or remove/qualify the X-Shooter-based generalization claim. The in-survey random-split res
  2. [Abstract; §5.2] The abstract and conclusions state that predictions for unseen spectral halves grow to 12–25% (5–15%) of the flux with distance from the cut-off. Section 5.2 shows only four individual examples and says the blue/red ends are too noisy for comparison. No aggregate P/O statistic, figure, or table analogous to Fig. 6 is provided to support this quantitative claim. Please add a quantitative evaluation of the half-mask test set, or revise the abstract/conclusions to match the qualitative evidence actually presented.
  3. [§4 vs §6.2] The BAL exclusion criterion is stated inconsistently. Section 4 says 'exclude any source with P_BAL > 0.0,' while Section 6.2 says 'aside from removing objects with P_BAL > 0.7 from training,' and the Fig. 12 caption states 'BALs were excluded from the training of AGNFormer.' These thresholds define materially different training sets (for example, whether mini-BALs with P_BAL=0.5/0.75 are included). Because Section 6.2 interprets the Civ overprediction as a consequence of BAL exclusion, please correct the threshold and specify the exact criterion in both places.
minor comments (6)
  1. [§3 vs App. B.1] The batch size is given as 16 in §3 but Appendix B.1 concludes 'we chose a batch size of 32.' With 1,000,000 training steps this is a factor-of-two difference in the number of spectra seen; please harmonize.
  2. [Fig. 8 caption] The caption refers to 'validation sources with P_BAL > 0.95,' while the text in §5.1 says the BAL subset is split out from the test sample. Please align the terminology.
  3. [§5.1 vs Conclusions] The full-sample broad-line dispersion is quoted as ≈10–15% in §5.1 but 10–16% in the abstract and conclusions; reconcile the numbers.
  4. [§5.1 vs Conclusions] A pull spread below 1 at mask edges is described in §5.1 as 'the uncertainties may be overestimated by ≈20% and the model is too conservative,' but the Conclusions say the model 'might be slightly overconfident.' If the pull width is <1, the model is underconfident, not overconfident; please fix the wording.
  5. [Abstract vs App. B.3] The abstract quotes an error floor of ≈2–6%, while Appendix B.3 reports individual floors from 1.6±0.2% (Hγ) to 6.5±0.3% (Hα). Align the stated range or describe it as approximate.
  6. [§6.1; §4.1] Please clarify the X-Shooter preprocessing: the only stated adaptation is interpolation to 4500 pixels, but §4.1 says SDSS spectra are not resampled. If this interpolation degrades X-Shooter's native resolution, the 'resolution invariance' claim should be qualified accordingly.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the masked-reconstruction targets are held-out pixels and the headline benchmark is externally defined.

full rationale

AGNFormer is trained with a masked-autoencoding objective in which pixels in the masked set M are zeroed and the negative log-likelihood (Eqs. 15-16) compares predictions against the observed fluxes f_i for those same pixels. The headline broad-line accuracies are measured on a disjoint 3% test split (70:27:3, Sec. 3), and Sec. 5.1 states that 'None of the spectra in our testing have been previously seen by the model.' Thus the reconstruction claim does not reduce by construction to the training objective. The external comparison uses the Greig et al. (2024) blind challenge, which the authors did not curate, and Table 2 reports parameter-free MAE values against eleven competing models. The only notable caveat is that the X-Shooter portion of that benchmark may involve objects whose SDSS spectra are inside DR16Q ('pretrained on the entire SDSS DR16Q catalogue'), which would weaken the cross-resolution generalization argument; however, that is a data-leakage/validity concern, not a circular-derivation concern: no equation or fitted parameter is being renamed as a prediction. The per-line 'error floors' in App. B.3 come from a two-parameter fit (sigma_model/f_pred = sqrt(a^2 + (b/S/N)^2)) used as a diagnostic of where prediction error dominates, not as a model output. Self-citations (e.g., Rozanski et al. 2024 and the 'in preparation' follow-ups) are incidental related-work references and are not load-bearing for the central claim. No circular step can be exhibited from the paper's own derivations.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claims rest on standard ML training on public data rather than on new physics, so the ledger is short: model/diagnostic hyperparameters and design choices (mask widths, normalization, masking mix), the domain assumptions that SDSS redshifts and P_BAL flags are correct, and the paper's own explicit Gaussian-independence simplification. No invented physical entities. The most consequential unstated premise is that a random split of SDSS DR16Q is a fair proxy for 'unseen' AGN spectra — this sets the scope of every headline percentage.

free parameters (5)
  • Per-region error-floor asymptote a = Oi 4.0±0.1%, Siiv 4.3±0.1%, Civ 4.6±0.1%, Ciii] 3.0±0.1%, Mgii 2.8±0.1%, Hγ 1.6±0.2%, Hβ 3.9±0.1%, Hα 6.5±0.3%
    Fitted in App. B.3 via σ_model/f_pred = sqrt(a² + (b/S/N)²); a diagnostic summary of the model noise floor, not used to derive the headline reconstruction claims.
  • Error-floor nuisance parameter b = not quoted
    Second parameter of the same B.3 fit; absorbs S/N-dependent noise. Diagnostic only.
  • Architecture hyperparameters (d_model=256, layers=8, heads=4, batch=32, 1M steps, 1 register token)
    Chosen via validation-loss experiments (App. B), standard ML practice; not fit to the benchmark data.
  • Mask design: ±10,000 km/s fixed mask width; 35:40:25 random-pixel:random-chunk:fixed masking split; Feii mask 4500-5300
    Design choices. The fixed-mask width defines what 'unseen broad line' means in every reported percentage; the masking mix shapes the learned representations.
  • Selection cuts: log10(L5100)>43.5; P_BAL exclusion threshold (0.0 or 0.7 — stated inconsistently) = 43.5; P_BAL > 0.0 (Sec. 4) vs > 0.7 (Sec. 6.2)
    Luminosity cut removes host-dominated sources; the P_BAL cut removes BALs from training. The inconsistent threshold statement affects interpretation of the BAL test in Sec. 6.2.
assumptions (4)
  • domain assumption SDSS DR16Q rest-frame wavelengths and quoted redshifts are accurate enough that spectral realignment and the ±10^4 km/s masks are correctly placed.
    Sec. 4.1: spectra are shifted using Wu & Shen (2022) redshifts, ~90% from the SDSS pipeline. Miscentered masks would bias P/O ratios near line peaks.
  • domain assumption P_BAL classifications (Lyke et al. 2020) reliably identify the BAL QSOs that were excluded from training.
    Sec. 4 footnote 2 notes P_BAL is tuned to HiBALs and less robust for LoBALs; the paper's own BAL overprediction residuals (Sec. 6.2, Fig. 8) show this assumption is imperfect.
  • ad hoc to paper The predictive distribution for each masked pixel is an independent Gaussian.
    Sec. 2.1.6 states this explicitly, noting it cannot capture multimodality (BAL presence/absence) or correlated uncertainty across a line profile; the pull-distribution calibration claims (Fig. 7) depend on it.
  • domain assumption A random 70:27:3 split of SDSS DR16Q is representative of 'unseen' AGN spectra.
    Sec. 3: test spectra share the survey's selection, noise, and calibration systematics with training, so the headline percentages are in-distribution metrics; the X-Shooter test is the only external check.

how reviews work

0 comments
Cite this review

Pith. "Pith review of AGNFormer I: Reconstruction of AGN spectra using a probabilistic transformer model." pith.science (2026). https://pith.science/paper/RP5PFWCU

@misc{pith2026260719640,
  author       = {Pith},
  title        = {Pith review of: AGNFormer I: Reconstruction of AGN spectra using a probabilistic transformer model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RP5PFWCU}},
  note         = {Machine review of arXiv:2607.19640}
}
abstract

We explore how an uncertainty-aware transformer-based architecture can leverage information embedded across the entire observed optical spectra of AGN, focusing on the algorithm's ability to predict unseen or masked parts of luminous AGN spectra. This provides a direct probe of the learnable correlations between AGN continua and broad lines. We introduce AGNFormer, a transformer model trained to predict the mean expected flux and variance in masked spectral regions (major broad lines to ${\pm}10^{4}$kms$^{-1}$; missing halves), inputting rest-frame spectral fluxes and uncertainties across the entire redshift range of the SDSS DR16 Quasar Catalogue. We evaluate the performance of the model on both full (no S/N limit) and high-quality (S/N > 10) spectral samples using the negative-log likelihood, and via comparisons with existing C IV and ly-a reconstruction algorithms. The model successfully reconstructs unseen AGN broad lines to better than 10-16% (4-8%) of the flux for the full (S/N > 10) test sets, up to an error floor of $\approx$2-6% of the flux at S/N $\approx$ 40, while predictions for larger unseen halves grow to 12-25% (5-15%) of the flux the further away they are from the cut-off wavelength of the seen input spectrum. Predictions faithfully reproduce the broad AGN spectral diversity across the entire optical and UV QSO main sequence parameter spaces, including both Gaussian and Lorentzian profile regimes, Feii complexes, and narrow emission lines. Performance is similar or better compared to previous spectral reconstruction algorithms. The high precision of the broad-line region reconstruction demonstrates that the method successfully aggregates information across the spectrum and highlights how the AGN continuum and weaker lines/complexes have the potential to assist astronomers in the extraction of the entire wealth of information embedded in AGN spectra.

Figures

Figures reproduced from arXiv: 2607.19640 by the authors.

Figure 1
Figure 1. Architecture of the probabilistic transformer used in this work. We apply a shared non-linear embedding to the fluxes and [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Distribution of redshift vs bolometric luminosity, [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Top: Median spectrum (per wavelength bin) from the en￾tire SDSS DR16Q, with the broad lines analysed in this work highlighted in grey. Bottom: Distribution of the rest-frame wave￾length coverage vs median S/N of the spectrum. Each of the ver￾tical lines denote interesting broad lines that we mask in testing. Sources with lower S/N (<10) and within the rest-frame wave￾length range of 1500–4000Å dominate amongst the f… view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Top: Optical-QSO Main Sequence (QMS; also known as EV1) showing RFe ii (EWFe ii/EWHβ) vs. FWHMHβ, adopting the same grid spacing as defined by Sulentic et al. (2002) to split the population. The line profiles of the sources that make up this plot can broadly be modeled…
Figure 5
Figure 5. Figure 5: Example DR16Q spectra (observed in black) and spectral reconstructions (mean prediction [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 7
Figure 7. Figure 7: Pull residual (r) distribution plots for each of the broad lines shown in [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]
Figure 8
Figure 8. Figure 8: The plot is identical to Fig [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]
Figure 9
Figure 9. Figure 9: Median predicted/observed ratios (denoted P/O) on the same scale as Fig.6 for the extended Hβ region for our test sam￾ple, split across bins of Hβ FWHM and RFe ii as highlighted in the QSO MS (top panel of [PITH_FULL_IMAGE:figures/full_fig_p011_9.png]
Figure 11
Figure 11. Figure 11: Comparison of the mean (top) and absolute error (bot￾tom) of the predicted/observed flux ratios from our transformer model (black) vs. 11 models (colours, as indicated among the four legends) from a challenge to predict 30 X-Shooter (left) and 30 SDSS (right) QSO spec…
Figure 12
Figure 12. Figure 12: Similar to Fig [PITH_FULL_IMAGE:figures/full_fig_p012_12.png]
Figure 13
Figure 13. Figure 13: Example predictions around the C iv line region from our transformer (red) and the MFICA model of Rankine et al. (2020) (green) for different levels and locations of absorption. when masking broad ranges (1300Å-1650Å), effectively draw￾ing on fainter BAL features pres…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

273 extracted references · 13 canonical work pages

  1. [1]

    D., Allende Prieto , C., et al

    Alam , S., Albareti , F. D., Allende Prieto , C., et al. 2015, , 219, 12

    2015

  2. [2]

    Alexander , D. M. & Hickox , R. C. 2012, , 56, 93

    2012

  3. [3]

    T., Hewett , P

    Allen , J. T., Hewett , P. C., Richardson , C. T., Ferland , G. J., & Baldwin , J. A. 2013, , 430, 3510

    2013

  4. [4]

    & Auli , M

    Baevski , A. & Auli , M. 2018, arXiv e-prints, arXiv:1809.10853

    arXiv 2018

  5. [5]

    2014, arXiv e-prints, arXiv:1409.0473

    Bahdanau , D., Cho , K., & Bengio , Y. 2014, arXiv e-prints, arXiv:1409.0473

    arXiv 2014

  6. [6]

    Baldwin , J. A. 1977, , 214, 679

    1977

  7. [7]

    Baldwin , J. A. 1997, in Astronomical Society of the Pacific Conference Series, Vol. 113, IAU Colloq. 159: Emission Lines in Active Galaxies: New Methods and Techniques, ed. B. M. Peterson , F.-Z. Cheng , & A. S. Wilson , 80

    1997

  8. [8]

    A., Burke , W

    Baldwin , J. A., Burke , W. L., Gaskell , C. M., & Wampler , E. J. 1978, , 273, 431

    1978

Show all 273 references
  1. [9]

    A., Phillips , M

    Baldwin , J. A., Phillips , M. M., & Terlevich , R. 1981, , 93, 5

    1981

  2. [10]

    & Laor , A

    Baskin , A. & Laor , A. 2004, , 350, L31

    2004

  3. [11]

    & Laor , A

    Baskin , A. & Laor , A. 2005, , 358, 1043

    2005

  4. [12]

    2013, , 432, 1525

    Baskin , A., Laor , A., & Hamann , F. 2013, , 432, 1525

    2013

  5. [13]

    2015, , 449, 1593

    Baskin , A., Laor , A., & Hamann , F. 2015, , 449, 1593

    2015

  6. [14]

    H., Fan , X., White , R

    Becker , R. H., Fan , X., White , R. L., et al. 2001, , 122, 2850

    2001

  7. [15]

    S., Schlegel , D

    Bolton , A. S., Schlegel , D. J., Aubourg , \'E ., et al. 2012, , 144, 144

    2012

  8. [16]

    Boroson , T. A. & Green , R. F. 1992, , 80, 109

    1992

  9. [17]

    2024, , 689, A289

    Cabrera-Vives , G., Moreno-Cartagena , D., Astorga , N., et al. 2024, , 689, A289

    2024

  10. [18]

    2025, , 699, A168

    C \'a diz-Leyton , M., Cabrera-Vives , G., Protopapas , P., et al. 2025, , 699, A168

    2025

  11. [19]

    2024, arXiv e-prints, arXiv:2410.18857

    Chun , S., Kim , W., Park , S., & Yun , S. 2024, arXiv e-prints, arXiv:2410.18857

    2024

  12. [20]

    C., Banerji , M., & Richards , G

    Coatman , L., Hewett , P. C., Banerji , M., & Richards , G. T. 2016, , 461, 647

    2016

  13. [21]

    2023, arXiv e-prints, arXiv:2309.16588

    Darcet , T., Oquab , M., Mairal , J., & Bojanowski , P. 2023, arXiv e-prints, arXiv:2309.16588

    2023 arXiv

  14. [22]

    2018, arXiv e-prints, arXiv:1810.04805

    Devlin , J., Chang , M.-W., Lee , K., & Toutanova , K. 2018, arXiv e-prints, arXiv:1810.04805

    2018 arXiv

  15. [23]

    C., et al

    Dietrich , M., Hamann , F., Shields , J. C., et al. 2002, , 581, 912

    2002

  16. [24]

    2023, , 670, A54

    Donoso-Oliva , C., Becker , I., Protopapas , P., et al. 2023, , 670, A54

    2023

  17. [25]

    2020, arXiv e-prints, arXiv:2010.11929

    Dosovitskiy , A., Beyer , L., Kolesnikov , A., et al. 2020, arXiv e-prints, arXiv:2010.11929

    2020 arXiv

  18. [26]

    N., Hall , P

    Filiz Ak , N., Brandt , W. N., Hall , P. B., et al. 2014, , 791, 88

    2014

  19. [27]

    B., Protopapas , P., Muthukrishna , D., & Brockmeier , A

    Fox Fortino , W., Bianco , F. B., Protopapas , P., Muthukrishna , D., & Brockmeier , A. 2025, arXiv e-prints, arXiv:2507.22106

    2025 arXiv

  20. [28]

    J., Hewett , P

    Francis , P. J., Hewett , P. C., Foltz , C. B., & Chaffee , F. H. 1992, , 398, 476

    1992

  21. [29]

    R., Jiang , L., Brandt , W

    Gibson , R. R., Jiang , L., Brandt , W. N., et al. 2009, , 692, 758

    2009

  22. [30]

    F., Schmidt , M., & Liebert , J

    Green , R. F., Schmidt , M., & Liebert , J. 1986, , 61, 305

    1986

  23. [31]

    Greig , B., Bosman , S. E. I., Davies , F. B., et al. 2024, arXiv e-prints, arXiv:2404.01556

    2024 arXiv

  24. [32]

    Gunn , J. E. & Peterson , B. A. 1965, , 142, 1633

    1965

  25. [33]

    Guo , C., Pleiss , G., Sun , Y., & Weinberger , K. Q. 2017, arXiv e-prints, arXiv:1706.04599

    2017 arXiv

  26. [34]

    1994, , 432, L95

    Haardt , F., Maraschi , L., & Ghisellini , G. 1994, , 432, L95

    1994

  27. [35]

    2021, arXiv e-prints, arXiv:2111.06377

    He , K., Chen , X., Xie , S., et al. 2021, arXiv e-prints, arXiv:2111.06377

    2021 arXiv

  28. [36]

    Heckman , T. M. 1980, , 88, 311

    1980

  29. [37]

    C., Filippenko , A

    Ho , L. C., Filippenko , A. V., & Sargent , W. L. 1995, , 98, 477

    1995

  30. [38]

    C., Filippenko , A

    Ho , L. C., Filippenko , A. V., & Sargent , W. L. W. 1997 a , , 112, 315

    1997

  31. [39]

    C., Filippenko , A

    Ho , L. C., Filippenko , A. V., & Sargent , W. L. W. 1997 b , , 487, 568

    1997

  32. [40]

    2024, arXiv e-prints, arXiv:2404.06395

    Hu , S., Tu , Y., Han , X., et al. 2024, arXiv e-prints, arXiv:2404.06395

    2024 arXiv

  33. [41]

    Jacobs, R. A. 1988, Neural Networks, 1, 295

    1988

  34. [42]

    M., Tremonti , C., et al

    Kauffmann , G., Heckman , T. M., Tremonti , C., et al. 2003, , 346, 1055

    2003

  35. [43]

    C., Bechtold , J., Trump , J

    Kelly , B. C., Bechtold , J., Trump , J. R., Vestergaard , M., & Siemiginowska , A. 2008, , 176, 355

    2008

  36. [44]

    J., Dopita , M

    Kewley , L. J., Dopita , M. A., Sutherland , R. S., Heisler , C. A., & Trevena , J. 2001, , 556, 121

    2001

  37. [45]

    J., Groves , B., Kauffmann , G., & Heckman , T

    Kewley , L. J., Groves , B., Kauffmann , G., & Heckman , T. 2006, , 372, 961

    2006

  38. [46]

    & Zetzl , M

    Kollatschny , W. & Zetzl , M. 2013, , 549, A100

    2013

  39. [47]

    2020, arXiv e-prints, arXiv:2008.10546

    Kong , L., Sun , J., & Zhang , C. 2020, arXiv e-prints, arXiv:2008.10546

    2020 arXiv

  40. [48]

    2020, arXiv e-prints, arXiv:2002.10118

    Kristiadi , A., Hein , M., & Hennig , P. 2020, arXiv e-prints, arXiv:2002.10118

    2020 arXiv

  41. [49]

    Krolik , J. H. & Begelman , M. C. 1988, , 329, 702

    1988

  42. [50]

    2016, arXiv e-prints, arXiv:1612.01474

    Lakshminarayanan , B., Pritzel , A., & Blundell , C. 2016, arXiv e-prints, arXiv:1612.01474

    2016 arXiv

  43. [51]

    2023, arXiv e-prints, arXiv:2310.03024

    Lanusse , F., Parker , L., Golkar , S., et al. 2023, arXiv e-prints, arXiv:2310.03024

    2023 arXiv

  44. [52]

    Leung , H. W. & Bovy , J. 2024, , 527, 1494

    2024

  45. [53]

    2019, arXiv e-prints, arXiv:1910.13461

    Lewis , M., Liu , Y., Goyal , N., et al. 2019, arXiv e-prints, arXiv:1910.13461

    2019 arXiv

  46. [54]

    2021, arXiv e-prints, arXiv:2103.13629

    Li , W., Huang , X., Lu , J., Feng , J., & Zhou , J. 2021, arXiv e-prints, arXiv:2103.13629

    2021 arXiv

  47. [55]

    L., et al

    L \'o pez , S., D'Odorico , V., Ellison , S. L., et al. 2016, , 594, A91

    2016

  48. [56]

    & Hutter , F

    Loshchilov , I. & Hutter , F. 2016, arXiv e-prints, arXiv:1608.03983

    2016 arXiv

  49. [57]

    & Hutter , F

    Loshchilov , I. & Hutter , F. 2017, arXiv e-prints, arXiv:1711.05101

    2017 arXiv

  50. [58]

    W., Higley , A

    Lyke , B. W., Higley , A. N., McLane , J. N., et al. 2020, , 250, 8

    2020

  51. [59]

    H., & Rieke , M

    Maiolino , R., Rieke , G. H., & Rieke , M. J. 1996, , 111, 537

    1996

  52. [60]

    2003, The Messenger, 114, 20

    Mayor , M., Pepe , F., Queloz , D., et al. 2003, The Messenger, 114, 20

    2003

  53. [61]

    2018, arXiv e-prints, arXiv:1812.06162

    McCandlish , S., Kaplan , J., Amodei , D., & OpenAI Dota Team . 2018, arXiv e-prints, arXiv:1812.06162

    2018 arXiv

  54. [62]

    Mittaz , J. P. D., Penston , M. V., & Snijders , M. A. J. 1990, , 242, 370

    1990

  55. [63]

    2014, arXiv e-prints, arXiv:1406.6247

    Mnih , V., Heess , N., Graves , A., & Kavukcuoglu , K. 2014, arXiv e-prints, arXiv:1406.6247

    2014 arXiv

  56. [64]

    & Ferland , G

    Mushotzky , R. & Ferland , G. J. 1984, , 278, 558

    1984

  57. [65]

    M., Nikutta , R., Ivezi \'c , Z ., & Elitzur , M

    Nenkova , M., Sirocky , M. M., Nikutta , R., Ivezi \'c , Z ., & Elitzur , M. 2008, , 685, 160

    2008

  58. [66]

    2014, arXiv e-prints, arXiv:1412.1897

    Nguyen , A., Yosinski , J., & Clune , J. 2014, arXiv e-prints, arXiv:1412.1897

    2014 arXiv

  59. [67]

    N., Tavker , S., & Efimov , D

    Oreshkin , B. N., Tavker , S., & Efimov , D. 2025, arXiv e-prints, arXiv:2508.16355

    2025 arXiv

  60. [68]

    2018, , 613, A51

    P \^a ris , I., Petitjean , P., Aubourg , \'E ., et al. 2018, , 613, A51

    2018

  61. [69]

    P., et al

    P \^a ris , I., Petitjean , P., Ross , N. P., et al. 2017, , 597, A79

    2017

  62. [70]

    Peterson , B. M. & Wandel , A. 1999, , 521, L95

    1999

  63. [71]

    Portillo , S. K. N., Parejko , J. K., Vergara , J. R., & Connolly , A. J. 2020, , 160, 45

    2020

  64. [72]

    & Ricci , C

    Ramos Almeida , C. & Ricci , C. 2017, Nature Astronomy, 1, 679

    2017

  65. [73]

    L., Hewett , P

    Rankine , A. L., Hewett , P. C., Banerji , M., & Richards , G. T. 2020, , 492, 4553

    2020

  66. [74]

    2026, arXiv e-prints, arXiv:2601.03753

    Rauba , P., Cikojevic , V., Bartolic , F., et al. 2026, arXiv e-prints, arXiv:2601.03753

    2026

  67. [75]

    1998, , 36, 267

    Rauch , M. 1998, , 36, 267

    1998

  68. [76]

    2020, , 898, L1

    Ricci , C., Kara , E., Loewenstein , M., et al. 2020, , 898, L1

    2020

  69. [77]

    T., Kruczek , N

    Richards , G. T., Kruczek , N. E., Gallagher , S. C., et al. 2011, , 141, 167

    2011

  70. [78]

    2024, arXiv e-prints, arXiv:2407.05751

    R \'o \.z a \'n ski , T., Ting , Y.-S., & Jab o \'n ska , M. 2024, arXiv e-prints, arXiv:2407.05751

    2024 arXiv

  71. [79]

    M., Schneider , P

    Rudolf , N., G \"u nther , H. M., Schneider , P. C., & Schmitt , J. H. M. M. 2016, , 585, A113

    2016

  72. [80]

    2025, arXiv e-prints, arXiv:2501.18965

    Schaipp , F., H \"a gele , A., Taylor , A., Simsekli , U., & Bach , F. 2025, arXiv e-prints, arXiv:2501.18965

    2025 arXiv

  73. [81]

    2012, arXiv e-prints, arXiv:1206.1106

    Schaul , T., Zhang , S., & LeCun , Y. 2012, arXiv e-prints, arXiv:1206.1106

    2012 arXiv

  74. [82]

    2023, , 526, 3037

    Scourfield , M., Saintonge , A., de Mijolla , D., & Viti , S. 2023, , 526, 3037

    2023

  75. [83]

    M., Kalmbach , J

    Sedaghat , N., Smart , B. M., Kalmbach , J. B., Howard , E. L., & Amindavar , H. 2023, , 526, 1559

    2023

  76. [84]

    Shakura , N. I. & Sunyaev , R. A. 1973, , 24, 337

    1973

  77. [85]

    2025, arXiv e-prints, arXiv:2506.02654

    Shen , S., Pan , B., & Xue , G. 2025, arXiv e-prints, arXiv:2506.02654

    2025 arXiv

  78. [86]

    N., Richards , G

    Shen , Y., Brandt , W. N., Richards , G. T., et al. 2016, , 831, 7

    2016

  79. [87]

    & Ho , L

    Shen , Y. & Ho , L. C. 2014, , 513, 210

    2014

  80. [88]

    C., et al

    Stepney , M., Banerji , M., Hewett , P. C., et al. 2023, , 524, 5497

    2023

  81. [89]

    W., Bachev , R., Marziani , P., Negrete , C

    Sulentic , J. W., Bachev , R., Marziani , P., Negrete , C. A., & Dultzin , D. 2007, , 666, 757

    2007

  82. [90]

    W., del Olmo , A., Marziani , P., et al

    Sulentic , J. W., del Olmo , A., Marziani , P., et al. 2017, , 608, A122

    2017

  83. [91]

    W., Marziani , P., Zamanov , R., et al

    Sulentic , J. W., Marziani , P., Zamanov , R., et al. 2002, , 566, L71

    2002

  84. [92]

    W., Zwitter , T., Marziani , P., & Dultzin-Hacyan , D

    Sulentic , J. W., Zwitter , T., Marziani , P., & Dultzin-Hacyan , D. 2000, , 536, L5

    2000

  85. [93]

    & Malkan , M

    Sun , W.-H. & Malkan , M. A. 1989, , 346, 68

    1989

  86. [94]

    H., & Tsvetanov , Z

    Tolea , A., Krolik , J. H., & Tsvetanov , Z. 2002, , 578, L31

    2002

  87. [95]

    2023, arXiv e-prints, arXiv:2307.09288

    Touvron , H., Martin , L., Stone , K., et al. 2023, arXiv e-prints, arXiv:2307.09288

    2023 arXiv

  88. [96]

    Tristram , K. R. W., Meisenheimer , K., Jaffe , W., et al. 2007, , 474, 837

    2007

  89. [97]

    D., Cushing , M

    Vacca , W. D., Cushing , M. C., & Rayner , J. T. 2003, , 115, 389

    2003

  90. [98]

    2017, arXiv e-prints, arXiv:1706.03762

    Vaswani , A., Shazeer , N., Parmar , N., et al. 2017, arXiv e-prints, arXiv:1706.03762

    2017 arXiv

  91. [99]

    & Fabian , A

    Vaughan , S. & Fabian , A. C. 2004, , 348, 1415

    2004

  92. [100]

    2011, , 536, A105

    Vernet , J., Dekker , H., D'Odorico , S., et al. 2011, , 536, A105

    2011

  93. [101]

    2024, arXiv e-prints, arXiv:2410.05192

    Wen , K., Li , Z., Wang , J., et al. 2024, arXiv e-prints, arXiv:2410.05192

    2024 arXiv

  94. [102]

    J., Carswell , R

    Weymann , R. J., Carswell , R. F., & Smith , M. G. 1981, , 19, 41

    1981

  95. [103]

    J., Morris , S

    Weymann , R. J., Morris , S. L., Foltz , C. B., & Hewett , P. C. 1991, , 373, 23

    1991

  96. [104]

    Wilkins , D. R. & Gallo , L. C. 2015, , 448, 703

    2015

  97. [105]

    & Tu , K

    Wu , H. & Tu , K. 2023, arXiv e-prints, arXiv:2311.15211

    2023 arXiv

  98. [106]

    & Shen , Y

    Wu , Q. & Shen , Y. 2022, , 263, 42

    2022

  99. [107]

    2020, arXiv e-prints, arXiv:2002.04745

    Xiong , R., Yang , Y., He , D., et al. 2020, arXiv e-prints, arXiv:2002.04745

    2020 arXiv

  100. [108]

    , keywords =

    The shape of broad-line profiles in active galactic nuclei. , keywords =. doi:10.1051/0004-6361/201219411 , archivePrefix =. 1211.3065 , primaryClass =

    arXiv

  101. [109]

    arXiv e-prints , keywords =

    Decoupled Weight Decay Regularization. arXiv e-prints , keywords =. doi:10.48550/arXiv.1711.05101 , archivePrefix =. 1711.05101 , primaryClass =

    doi arXiv

  102. [110]

    arXiv e-prints , keywords =

    Learning Probabilistic Ordinal Embeddings for Uncertainty-Aware Regression. arXiv e-prints , keywords =. doi:10.48550/arXiv.2103.13629 , archivePrefix =. 2103.13629 , primaryClass =

    doi arXiv

  103. [111]

    , keywords =

    Uncertainty estimation for time series classification: Exploring predictive uncertainty in transformer-based models for variable stars. , keywords =. doi:10.1051/0004-6361/202453388 , archivePrefix =. 2412.10528 , primaryClass =

    arXiv

  104. [112]

    arXiv e-prints , keywords =

    Probabilistic Pretraining for Neural Regression. arXiv e-prints , keywords =. doi:10.48550/arXiv.2508.16355 , archivePrefix =. 2508.16355 , primaryClass =

    doi arXiv

  105. [113]

    arXiv e-prints , keywords =

    Probabilistic Transformers for Joint Modeling of Global Weather Dynamics and Decision-Centric Variables. arXiv e-prints , keywords =. doi:10.48550/arXiv.2601.03753 , archivePrefix =. 2601.03753 , primaryClass =

    doi

  106. [114]

    , keywords =

    What controls the [OIII] 5007 line strength in active galactic nuclei?. , keywords =. doi:10.1111/j.1365-2966.2005.08841.x , archivePrefix =. astro-ph/0501436 , primaryClass =

    2005

  107. [115]

    , keywords =

    What does CIV 1549 tell us about the physical driver of the Eigenvector quasar sequence?. , keywords =. doi:10.1051/0004-6361/201630309 , archivePrefix =. 1708.03187 , primaryClass =

    arXiv

  108. [116]

    , keywords =

    C IV emission-line properties and systematic trends in quasar black hole mass estimates. , keywords =. doi:10.1093/mnras/stw1360 , archivePrefix =. 1606.02726 , primaryClass =

    arXiv

  109. [117]

    arXiv e-prints , keywords =

    Probabilistic Transformer: A Probabilistic Dependency Model for Contextual Word Representation. arXiv e-prints , keywords =. doi:10.48550/arXiv.2311.15211 , archivePrefix =. 2311.15211 , primaryClass =

    doi arXiv

  110. [118]

    arXiv e-prints , keywords =

    A Pretrained Probabilistic Transformer for City-Scale Traffic Volume Prediction. arXiv e-prints , keywords =. doi:10.48550/arXiv.2506.02654 , archivePrefix =. 2506.02654 , primaryClass =

    doi arXiv

  111. [119]

    arXiv e-prints , keywords =

    Probabilistic Language-Image Pre-Training. arXiv e-prints , keywords =. doi:10.48550/arXiv.2410.18857 , archivePrefix =. 2410.18857 , primaryClass =

    doi

  112. [120]

    arXiv e-prints , keywords =

    SDE-Net: Equipping Deep Neural Networks with Uncertainty Estimates. arXiv e-prints , keywords =. doi:10.48550/arXiv.2008.10546 , archivePrefix =. 2008.10546 , primaryClass =

    2008 doi arXiv

  113. [121]

    arXiv e-prints , keywords =

    On Calibration of Modern Neural Networks. arXiv e-prints , keywords =. doi:10.48550/arXiv.1706.04599 , archivePrefix =. 1706.04599 , primaryClass =

    doi arXiv

  114. [122]

    arXiv e-prints , keywords =

    Deep Neural Networks are Easily Fooled: High Confidence Predictions for Unrecognizable Images. arXiv e-prints , keywords =. doi:10.48550/arXiv.1412.1897 , archivePrefix =. 1412.1897 , primaryClass =

    doi arXiv

  115. [123]

    arXiv e-prints , keywords =

    Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles. arXiv e-prints , keywords =. doi:10.48550/arXiv.1612.01474 , archivePrefix =. 1612.01474 , primaryClass =

    doi arXiv

  116. [124]

    arXiv e-prints , keywords =

    Being Bayesian, Even Just a Bit, Fixes Overconfidence in ReLU Networks. arXiv e-prints , keywords =. doi:10.48550/arXiv.2002.10118 , archivePrefix =. 2002.10118 , primaryClass =

    2002 doi arXiv

  117. [125]

    arXiv e-prints , keywords =

    Open Problems in Mechanistic Interpretability. arXiv e-prints , keywords =. doi:10.48550/arXiv.2501.16496 , archivePrefix =. 2501.16496 , primaryClass =

    doi arXiv

  118. [126]

    arXiv e-prints , keywords =

    Neural Machine Translation by Jointly Learning to Align and Translate. arXiv e-prints , keywords =. doi:10.48550/arXiv.1409.0473 , archivePrefix =. 1409.0473 , primaryClass =

    doi arXiv

  119. [127]

    arXiv e-prints , keywords =

    BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension. arXiv e-prints , keywords =. doi:10.48550/arXiv.1910.13461 , archivePrefix =. 1910.13461 , primaryClass =

    1910 doi arXiv

  120. [128]

    arXiv e-prints , keywords =

    Masked Autoencoders Are Scalable Vision Learners. arXiv e-prints , keywords =. doi:10.48550/arXiv.2111.06377 , archivePrefix =. 2111.06377 , primaryClass =

    doi arXiv

  121. [129]

    arXiv e-prints , keywords =

    ABC-SN: Attention Based Classifier for Supernova Spectra. arXiv e-prints , keywords =. doi:10.48550/arXiv.2507.22106 , archivePrefix =. 2507.22106 , primaryClass =

    doi arXiv

  122. [130]

    arXiv e-prints , keywords =

    Layer Normalization. arXiv e-prints , keywords =. doi:10.48550/arXiv.1607.06450 , archivePrefix =. 1607.06450 , primaryClass =

    doi arXiv

  123. [131]

    , keywords =

    ATAT: Astronomical Transformer for time series and Tabular data. , keywords =. doi:10.1051/0004-6361/202449475 , archivePrefix =. 2405.03078 , primaryClass =

    arXiv

  124. [132]

    , keywords =

    An Objective Classification Scheme for QSO Spectra. , keywords =. doi:10.1086/171870 , adsurl =

    doi

  125. [133]

    , keywords =

    A strong redshift dependence of the broad absorption line quasar fraction. , keywords =. doi:10.1111/j.1365-2966.2010.17489.x , archivePrefix =. 1007.3991 , primaryClass =

    2010

  126. [134]

    , keywords =

    Quasar Factor Analysis-An Unsupervised and Probabilistic Quasar Continuum Prediction Algorithm with Latent Factor Analysis. , keywords =. doi:10.3847/1538-4365/acf2f1 , archivePrefix =. 2211.11784 , primaryClass =

    arXiv

  127. [135]

    arXiv e-prints , keywords =

    BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. arXiv e-prints , keywords =. doi:10.48550/arXiv.1810.04805 , archivePrefix =. 1810.04805 , primaryClass =

    doi arXiv

  128. [136]

    , keywords =

    The Dependence of C IV Broad Absorption Line Properties on Accompanying Si IV and Al III Absorption: Relating Quasar-wind Ionization Levels, Kinematics, and Column Densities. , keywords =. doi:10.1088/0004-637X/791/2/88 , archivePrefix =. 1407.2250 , primaryClass =

    arXiv

  129. [137]

    arXiv e-prints , keywords =

    An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale. arXiv e-prints , keywords =. doi:10.48550/arXiv.2010.11929 , archivePrefix =. 2010.11929 , primaryClass =

    2010 doi arXiv

  130. [138]

    , keywords =

    Broad Absorption Line Quasars in the Early Data Release from the Sloan Digital Sky Survey. , keywords =. doi:10.1086/344563 , archivePrefix =. astro-ph/0209081 , primaryClass =

    arXiv

  131. [139]

    , keywords =

    A Catalog of Broad Absorption Line Quasars in Sloan Digital Sky Survey Data Release 5. , keywords =. doi:10.1088/0004-637X/692/1/758 , archivePrefix =. 0810.2747 , primaryClass =

    arXiv

  132. [140]

    arXiv e-prints , keywords =

    Attention, Please! Revisiting Attentive Probing for Masked Image Modeling. arXiv e-prints , keywords =. doi:10.48550/arXiv.2506.10178 , archivePrefix =. 2506.10178 , primaryClass =

    doi

  133. [141]

    arXiv e-prints , keywords =

    CA-Stream: Attention-based pooling for interpretable image recognition. arXiv e-prints , keywords =. doi:10.48550/arXiv.2404.14996 , archivePrefix =. 2404.14996 , primaryClass =

    doi arXiv

  134. [142]

    arXiv e-prints , keywords =

    Self-Attentive Pooling for Efficient Deep Learning. arXiv e-prints , keywords =. doi:10.48550/arXiv.2209.07659 , archivePrefix =. 2209.07659 , primaryClass =

    doi arXiv

  135. [143]

    arXiv e-prints , keywords =

    Context-aware Attentional Pooling (CAP) for Fine-grained Visual Classification. arXiv e-prints , keywords =. doi:10.48550/arXiv.2101.06635 , archivePrefix =. 2101.06635 , primaryClass =

    doi arXiv

  136. [145]

    arXiv e-prints , keywords =

    On Layer Normalization in the Transformer Architecture. arXiv e-prints , keywords =. doi:10.48550/arXiv.2002.04745 , archivePrefix =. 2002.04745 , primaryClass =

    2002 doi arXiv

  137. [146]

    arXiv e-prints , keywords =

    Adaptive Input Representations for Neural Language Modeling. arXiv e-prints , keywords =. doi:10.48550/arXiv.1809.10853 , archivePrefix =. 1809.10853 , primaryClass =

    doi arXiv

  138. [147]

    arXiv e-prints , keywords =

    Vision Transformers Need Registers. arXiv e-prints , keywords =. doi:10.48550/arXiv.2309.16588 , archivePrefix =. 2309.16588 , primaryClass =

    doi arXiv

  139. [148]

    arXiv e-prints , keywords =

    Understanding intermediate layers using linear classifier probes. arXiv e-prints , keywords =. doi:10.48550/arXiv.1610.01644 , archivePrefix =. 1610.01644 , primaryClass =

    doi arXiv

  140. [149]

    arXiv e-prints , keywords =

    Perception Encoder: The best visual embeddings are not at the output of the network. arXiv e-prints , keywords =. doi:10.48550/arXiv.2504.13181 , archivePrefix =. 2504.13181 , primaryClass =

    doi arXiv

  141. [150]

    arXiv e-prints , keywords =

    Llama 2: Open Foundation and Fine-Tuned Chat Models. arXiv e-prints , keywords =. doi:10.48550/arXiv.2307.09288 , archivePrefix =. 2307.09288 , primaryClass =

    doi arXiv

  142. [151]

    arXiv e-prints , keywords =

    SGDR: Stochastic Gradient Descent with Warm Restarts. arXiv e-prints , keywords =. doi:10.48550/arXiv.1608.03983 , archivePrefix =. 1608.03983 , primaryClass =

    doi arXiv

  143. [152]

    arXiv e-prints , keywords =

    MiniCPM: Unveiling the Potential of Small Language Models with Scalable Training Strategies. arXiv e-prints , keywords =. doi:10.48550/arXiv.2404.06395 , archivePrefix =. 2404.06395 , primaryClass =

    doi arXiv

  144. [153]

    arXiv e-prints , keywords =

    Understanding Warmup-Stable-Decay Learning Rates: A River Valley Loss Landscape Perspective. arXiv e-prints , keywords =. doi:10.48550/arXiv.2410.05192 , archivePrefix =. 2410.05192 , primaryClass =

    doi arXiv

  145. [154]

    arXiv e-prints , keywords =

    The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training. arXiv e-prints , keywords =. doi:10.48550/arXiv.2501.18965 , archivePrefix =. 2501.18965 , primaryClass =

    doi arXiv

  146. [155]

    arXiv e-prints , keywords =

    No More Pesky Learning Rates. arXiv e-prints , keywords =. doi:10.48550/arXiv.1206.1106 , archivePrefix =. 1206.1106 , primaryClass =

    doi arXiv

  147. [156]

    Jacobs , abstract =

    Robert A. Jacobs , abstract =. Increased rates of convergence through learning rate adaptation , journal =. 1988 , issn =. doi:https://doi.org/10.1016/0893-6080(88)90003-2 , url =

    1988 doi

  148. [157]

    , keywords =

    The Palomar-Green Catalog of Ultraviolet-Excess Stellar Objects. , keywords =. doi:10.1086/191115 , adsurl =

    doi

  149. [158]

    arXiv e-prints , keywords =

    An Empirical Model of Large-Batch Training. arXiv e-prints , keywords =. doi:10.48550/arXiv.1812.06162 , archivePrefix =. 1812.06162 , primaryClass =

    doi arXiv

  150. [159]

    , keywords =

    No redshift evolution in the rest-frame ultraviolet emission line properties of quasars from z = 1.5 to z = 4.0. , keywords =. doi:10.1093/mnras/stad2060 , archivePrefix =. 2307.02962 , primaryClass =

    arXiv

  151. [160]

    , keywords =

    The average absorption properties of broad absorption line quasars at 800 < _ rest < 3000 A , and the underlying physical parameters. , keywords =. doi:10.1093/mnras/stt582 , archivePrefix =. 1304.0986 , primaryClass =

    arXiv

  152. [161]

    , keywords =

    Absorption lines in the spectra of Quasi-Stellar Objects. , keywords =. doi:10.1146/annurev.aa.19.090181.000353 , adsurl =

  153. [162]

    , keywords =

    On the origins of C IV absorption profile diversity in broad absorption line quasars. , keywords =. doi:10.1093/mnras/stv406 , archivePrefix =. 1502.06609 , primaryClass =

    arXiv

  154. [163]

    arXiv e-prints , keywords =

    Quantifying Attention Flow in Transformers. arXiv e-prints , keywords =. doi:10.48550/arXiv.2005.00928 , archivePrefix =. 2005.00928 , primaryClass =

    2005 doi arXiv

  155. [164]

    arXiv e-prints , keywords =

    Learning to Deceive with Attention-Based Explanations. arXiv e-prints , keywords =. doi:10.48550/arXiv.1909.07913 , archivePrefix =. 1909.07913 , primaryClass =

    1909 doi arXiv

  156. [165]

    , keywords =

    The Discovery of a Luminous Broad Absorption Line Quasar at a Redshift of 7.02. , keywords =. doi:10.3847/2041-8213/aaf1d2 , archivePrefix =. 1810.11925 , primaryClass =

    arXiv

  157. [166]

    , keywords =

    The Multi-object, Fiber-fed Spectrographs for the Sloan Digital Sky Survey and the Baryon Oscillation Spectroscopic Survey. , keywords =. doi:10.1088/0004-6256/146/2/32 , archivePrefix =. 1208.2233 , primaryClass =

    arXiv

  158. [167]

    , keywords =

    Unification of Luminous Type 1 Quasars through C IV Emission. , keywords =. doi:10.1088/0004-6256/141/5/167 , archivePrefix =. 1011.2282 , primaryClass =

    arXiv

  159. [168]

    , keywords =

    XQ-100: A legacy survey of one hundred 3.5 z 4.5 quasars observed with VLT/X-shooter. , keywords =. doi:10.1051/0004-6361/201628161 , archivePrefix =. 1607.08776 , primaryClass =

    arXiv

  160. [169]

    , keywords =

    X-shooter, the new wide band intermediate resolution spectrograph at the ESO Very Large Telescope. , keywords =. doi:10.1051/0004-6361/201117752 , archivePrefix =. 1110.1944 , primaryClass =

    1944 arXiv

  161. [170]

    , keywords =

    Spectral Properties of Quasars from Sloan Digital Sky Survey Data Release 14: The Catalog. , keywords =. doi:10.3847/1538-4365/ab99c5 , archivePrefix =. 1910.10395 , primaryClass =

    1910 arXiv

  162. [171]

    , keywords =

    Classification parameters for the emission-line spectra of extragalactic objects. , keywords =. doi:10.1086/130766 , adsurl =

    doi

  163. [172]

    , keywords =

    Keplerian Motion of Broad-Line Region Gas as Evidence for Supermassive Black Holes in Active Galactic Nuclei. , keywords =. doi:10.1086/312190 , archivePrefix =. astro-ph/9905382 , primaryClass =

    arXiv

  164. [173]

    , keywords =

    The analysis of the broad hydrogen Balmer line ratios: Possible implications for the physical properties of the broad line region of AGNs. , keywords =. doi:10.1051/0004-6361/201219299 , archivePrefix =. 1205.3950 , primaryClass =

    arXiv

  165. [174]

    , keywords =

    The Lyman Alpha Forest in the Spectra of QSOs. , keywords =. doi:10.1146/annurev.astro.36.1.267 , archivePrefix =. astro-ph/9806286 , primaryClass =

    arXiv

  166. [175]

    , keywords =

    Phenomenology of Broad Emission Lines in Active Galactic Nuclei. , keywords =. doi:10.1146/annurev.astro.38.1.521 , adsurl =

    doi

  167. [176]

    , keywords =

    The host galaxies and classification of active galactic nuclei. , keywords =. doi:10.1111/j.1365-2966.2006.10859.x , archivePrefix =. astro-ph/0605681 , primaryClass =

    2006

  168. [177]

    , keywords =

    Theoretical Modeling of Starburst Galaxies. , keywords =. doi:10.1086/321545 , archivePrefix =. astro-ph/0106324 , primaryClass =

    arXiv

  169. [178]

    , keywords =

    The host galaxies of active galactic nuclei. , keywords =. doi:10.1111/j.1365-2966.2003.07154.x , archivePrefix =. astro-ph/0304239 , primaryClass =

    2003

  170. [179]

    The Sloan Digital Sky Survey Quasar Catalog. V. Seventh Data Release. , keywords =. doi:10.1088/0004-6256/139/6/2360 , archivePrefix =. 1004.1167 , primaryClass =

    arXiv

  171. [180]

    Central Masses and Broad-Line Region Sizes of Active Galactic Nuclei. II. A Homogeneous Analysis of a Large Reverberation-Mapping Database. , keywords =. doi:10.1086/423269 , archivePrefix =. astro-ph/0407299 , primaryClass =

    arXiv

  172. [181]

    , keywords =

    The Sloan Digital Sky Survey Reverberation Mapping Project: Estimating Masses of Black Holes in Quasars with Single-epoch Spectroscopy. , keywords =. doi:10.3847/1538-4357/abbc1c , archivePrefix =. 2007.02963 , primaryClass =

    2007 arXiv

  173. [182]

    , year = 1974, month = may, volume =

    The morphology of extragalactic radio sources of high and low luminosity. , year = 1974, month = may, volume =. doi:10.1093/mnras/167.1.31P , adsurl =

    1974 doi

  174. [183]

    , keywords =

    Resolving the complex structure of the dust torus in the active nucleus of the Circinus galaxy. , keywords =. doi:10.1051/0004-6361:20078369 , archivePrefix =. 0709.0209 , primaryClass =

    arXiv

  175. [184]

    , keywords =

    Molecular Tori in Seyfert Galaxies: Feeding the Monster and Hiding It. , keywords =. doi:10.1086/166414 , adsurl =

    doi

  176. [185]

    AGN Dusty Tori. II. Observational Implications of Clumpiness. , keywords =. doi:10.1086/590483 , archivePrefix =. 0806.0512 , primaryClass =

    arXiv

  177. [186]

    A Search for ``Dwarf'' Seyfert Nuclei. II. an Optical Spectral Atlas of the Nuclei of Nearby Galaxies. , keywords =. doi:10.1086/192170 , adsurl =

    doi

  178. [187]

    A Search for ``Dwarf'' Seyfert Nuclei. III. Spectroscopic Parameters and Properties of the Host Galaxies. , keywords =. doi:10.1086/313041 , archivePrefix =. astro-ph/9704107 , primaryClass =

    arXiv

  179. [188]

    A Search for ``Dwarf'' Seyfert Nuclei. V. Demographics of Nuclear Activity in Nearby Galaxies. , keywords =. doi:10.1086/304638 , archivePrefix =. astro-ph/9704108 , primaryClass =

    arXiv

  180. [189]

    , keywords =

    Fitting Improved Accretion Disk Models to the Multiwavelength Continua of Quasars and Active Galactic Nuclei. , keywords =. doi:10.1086/167986 , adsurl =

    doi

  181. [190]

    , keywords =

    What drives the growth of black holes?. , keywords =. doi:10.1016/j.newar.2011.11.003 , archivePrefix =. 1112.1949 , primaryClass =

    2011 arXiv

  182. [191]

    , keywords =

    Observational Constraints on the Dependence of Radio-Quiet Quasar X-Ray Emission on Black Hole Mass and Accretion Rate. , keywords =. doi:10.1086/533440 , archivePrefix =. 0801.2383 , primaryClass =

    arXiv

  183. [192]

    Observational appearance

    Black holes in binary systems. Observational appearance. , year = 1973, month = jan, volume =

    1973

  184. [193]

    , keywords =

    The Destruction and Recreation of the X-Ray Corona in a Changing-look Active Galactic Nucleus. , keywords =. doi:10.3847/2041-8213/ab91a1 , archivePrefix =. 2007.07275 , primaryClass =

    arXiv

  185. [194]

    , keywords =

    The Comptonization of accretion disc X-ray emission: consequences for X-ray reflection and the geometry of AGN coronae. , keywords =. doi:10.1093/mnras/stu2524 , archivePrefix =. 1412.0015 , primaryClass =

    arXiv

  186. [195]

    , keywords =

    A Model for the X-Ray and Ultraviolet Emission from Seyfert Galaxies and Galactic Black Holes. , keywords =. doi:10.1086/187520 , archivePrefix =. astro-ph/9405059 , primaryClass =

    arXiv

  187. [196]

    Nature Astronomy , keywords =

    Nuclear obscuration in active galactic nuclei. Nature Astronomy , keywords =. doi:10.1038/s41550-017-0232-z , archivePrefix =. 1709.00019 , primaryClass =

    arXiv

  188. [197]

    , keywords =

    Coronal-Line Forest AGN: the best view of the inner edge of the AGN torus?. , keywords =. doi:10.1093/mnras/stv113 , archivePrefix =. 1501.02705 , primaryClass =

    arXiv

  189. [198]

    Black hole mass estimation by means of broad emission lines

    Active galactic nuclei at z 1.5 - II. Black hole mass estimation by means of broad emission lines. , keywords =. doi:10.1093/mnras/stw568 , archivePrefix =. 1603.03437 , primaryClass =

    arXiv

  190. [199]

    Spectral energy distribution and accretion discs

    Active galactic nuclei at z 1.5 - I. Spectral energy distribution and accretion discs. , keywords =. doi:10.1093/mnras/stu2266 , archivePrefix =. 1410.8137 , primaryClass =

    arXiv

  191. [200]

    , keywords =

    Estimating black hole masses: Accretion disk fitting versus reverberation mapping and single epoch. , keywords =. doi:10.1051/0004-6361/201936218 , archivePrefix =. 1907.00986 , primaryClass =

    1907 arXiv

  192. [201]

    arXiv e-prints , keywords =

    TransformerPayne: enhancing spectral emulation accuracy and data efficiency by capturing long-range correlations. arXiv e-prints , keywords =. doi:10.48550/arXiv.2407.05751 , archivePrefix =. 2407.05751 , primaryClass =

    doi arXiv

  193. [202]

    , keywords =

    Identification of multicomponent LOFAR sources with multimodal deep learning. , keywords =. doi:10.1093/mnras/stae1454 , archivePrefix =. 2405.18584 , primaryClass =

    arXiv

  194. [204]

    , keywords =

    Ultraviolet variability of NGC 4151 : a study using principal component analysis. , keywords =. doi:10.1093/mnras/242.3.370 , adsurl =

    doi

  195. [205]

    arXiv e-prints , keywords =

    Machine Learning in Astronomy: a practical overview. arXiv e-prints , keywords =. doi:10.48550/arXiv.1904.07248 , archivePrefix =. 1904.07248 , primaryClass =

    1904 doi arXiv

  196. [206]

    , keywords =

    Ly emission-line reconstruction for high-z QSOs. , keywords =. doi:10.1093/mnras/stw3210 , archivePrefix =. 1605.09388 , primaryClass =

    arXiv

  197. [207]

    , keywords =

    Influence of the optical Fe II quasi-continuum on measuring the spectral parameters of active galactic nuclei. , keywords =. doi:10.1051/0004-6361/202347043 , archivePrefix =. 2309.14852 , primaryClass =

    arXiv

  198. [208]

    , keywords =

    Analysis of Optical Fe II Emission in a Sample of Active Galactic Nucleus Spectra. , keywords =. doi:10.1088/0067-0049/189/1/15 , archivePrefix =. 1004.2212 , primaryClass =

    arXiv

  199. [209]

    arXiv e-prints , keywords =

    Generating astronomical spectra from photometry with conditional diffusion models. arXiv e-prints , keywords =. doi:10.48550/arXiv.2211.05556 , archivePrefix =. 2211.05556 , primaryClass =

    doi arXiv

  200. [210]

    , keywords =

    De-noising of galaxy optical spectra with autoencoders. , keywords =. doi:10.1093/mnras/stad2709 , archivePrefix =. 2309.02315 , primaryClass =

    arXiv

  201. [211]

    , keywords =

    Dimensionality Reduction of SDSS Spectra with Variational Autoencoders. , keywords =. doi:10.3847/1538-3881/ab9644 , archivePrefix =. 2002.10464 , primaryClass =

    2002 arXiv

  202. [212]

    IAU Colloq

    Broad Emission Lines in Active Galactic Nuclei. IAU Colloq. 159: Emission Lines in Active Galaxies: New Methods and Techniques , year = 1997, editor =

    1997

  203. [213]

    arXiv e-prints , keywords =

    RoFormer: Enhanced Transformer with Rotary Position Embedding. arXiv e-prints , keywords =. doi:10.48550/arXiv.2104.09864 , archivePrefix =. 2104.09864 , primaryClass =

    doi arXiv

  204. [214]

    International Conference on Learning Representations , year=

    Are Transformers universal approximators of sequence-to-sequence functions? , author=. International Conference on Learning Representations , year=

  205. [215]

    , keywords =

    On the origin of the C IV Baldwin effect in active galactic nuclei. , keywords =. doi:10.1111/j.1365-2966.2004.07833.x , archivePrefix =. astro-ph/0403365 , primaryClass =

    2004

  206. [216]

    Determining Central Black Hole Masses in Distant Active Galaxies and Quasars. II. Improved Optical and UV Scaling Relationships. , keywords =. doi:10.1086/500572 , archivePrefix =. astro-ph/0601303 , primaryClass =

    arXiv

  207. [217]

    , keywords =

    Continuum and Emission-Line Strength Relations for a Large Active Galactic Nuclei Sample. , keywords =. doi:10.1086/344410 , archivePrefix =. astro-ph/0208348 , primaryClass =

    arXiv

  208. [218]

    , year = 1977, month = jun, volume =

    Luminosity Indicators in the Spectra of Quasi-Stellar Objects. , year = 1977, month = jun, volume =. doi:10.1086/155294 , adsurl =

    1977 doi

  209. [219]

    , keywords =

    The Low-luminosity End of the Radius-Luminosity Relationship for Active Galactic Nuclei. , keywords =. doi:10.1088/0004-637X/767/2/149 , archivePrefix =. 1303.1742 , primaryClass =

    arXiv

  210. [220]

    Open Astronomy , keywords =

    Broad spectral lines in AGNs and supermassive black hole mass measurements. Open Astronomy , keywords =. doi:10.1515/astro-2020-0003 , archivePrefix =. 1912.10502 , primaryClass =

    2020 arXiv

  211. [221]

    arXiv e-prints , keywords =

    AstroCLIP: Cross-Modal Pre-Training for Astronomical Foundation Models. arXiv e-prints , keywords =. doi:10.48550/arXiv.2310.03024 , archivePrefix =. 2310.03024 , primaryClass =

    doi arXiv

  212. [222]

    , keywords =

    Spectral Classification and Redshift Measurement for the SDSS-III Baryon Oscillation Spectroscopic Survey. , keywords =. doi:10.1088/0004-6256/144/5/144 , archivePrefix =. 1207.7326 , primaryClass =

    arXiv

  213. [223]

    , keywords =

    Comparing Single-epoch Virial Black Hole Mass Estimators for Luminous Quasars. , keywords =. doi:10.1088/0004-637X/753/2/125 , archivePrefix =. 1203.0601 , primaryClass =

    arXiv

  214. [224]

    , keywords =

    The Sloan Digital Sky Survey Reverberation Mapping Project: Velocity Shifts of Quasar Emission Lines. , keywords =. doi:10.3847/0004-637X/831/1/7 , archivePrefix =. 1602.03894 , primaryClass =

    arXiv

  215. [225]

    , keywords =

    Estimating Black Hole Masses in Active Galaxies Using the H Emission Line. , keywords =. doi:10.1086/431897 , archivePrefix =. astro-ph/0508335 , primaryClass =

    arXiv

  216. [226]

    Bulletin of the Astronomical Society of India , keywords =

    The mass of quasars. Bulletin of the Astronomical Society of India , keywords =. doi:10.48550/arXiv.1302.2643 , archivePrefix =. 1302.2643 , primaryClass =

    doi arXiv

  217. [227]

    , keywords =

    Stellar Karaoke: deep blind separation of terrestrial atmospheric effects out of stellar spectra by velocity whitening. , keywords =. doi:10.1093/mnras/stad2686 , archivePrefix =. 2301.00313 , primaryClass =

    arXiv

  218. [228]

    , keywords =

    Correction of the Atmospheric Transmission in Infrared Spectroscopy. , keywords =. doi:10.1086/117804 , adsurl =

    doi

  219. [229]

    , keywords =

    A Method of Correcting Near-Infrared Spectra for Telluric Absorption. , keywords =. doi:10.1086/346193 , archivePrefix =. astro-ph/0211255 , primaryClass =

    arXiv

  220. [230]

    , keywords =

    Modelling telluric line spectra in the optical and infrared with an application to VLT/X-Shooter spectra. , keywords =. doi:10.1051/0004-6361/201322749 , archivePrefix =. 1511.04641 , primaryClass =

    arXiv

  221. [231]

    The Messenger , year = 2003, month = dec, volume =

    Setting New Standards with HARPS. The Messenger , year = 2003, month = dec, volume =

    2003

  222. [232]

    , keywords =

    Relative quasar luminosities determined from emission line strenghts. , keywords =. doi:10.1038/273431a0 , adsurl =

    doi

  223. [233]

    , keywords =

    The line continuum luminosity ratio in active galactic nuclei : on the``Balwin effect''. , keywords =. doi:10.1086/161822 , adsurl =

    doi

  224. [234]

    , keywords =

    The broad line region in active nuclei and quasars - Correlations with luminosity and radio emission. , keywords =

  225. [235]

    Characterizing Extreme Emission-line Galaxies. I. A Four-zone Ionization Model for Very High-ionization Emission. , keywords =. doi:10.3847/1538-4357/ac141b , archivePrefix =. 2105.12765 , primaryClass =

    arXiv

  226. [236]

    , keywords =

    Evidence for Reionization at z -0.5ex 6: Detection of a Gunn-Peterson Trough in a z=6.28 Quasar. , keywords =. doi:10.1086/324231 , archivePrefix =. astro-ph/0108097 , primaryClass =

    arXiv

  227. [237]

    , year = 1965, month = nov, volume =

    On the Density of Neutral Hydrogen in Intergalactic Space. , year = 1965, month = nov, volume =. doi:10.1086/148444 , adsurl =

    1965 doi

  228. [238]

    , keywords =

    Machine learning for transient discovery in Pan-STARRS1 difference imaging. , keywords =. doi:10.1093/mnras/stv292 , archivePrefix =. 1501.05470 , primaryClass =

    arXiv

  229. [239]

    , keywords =

    CPz: Classification-aided photometric-redshift estimation. , keywords =. doi:10.1051/0004-6361/201730763 , archivePrefix =. 1808.04977 , primaryClass =

    arXiv

  230. [240]

    , keywords =

    Automated Classification of Periodic Variable Stars Detected by the Wide-field Infrared Survey Explorer. , keywords =. doi:10.1088/0004-6256/148/1/21 , archivePrefix =. 1402.0125 , primaryClass =

    arXiv

  231. [241]

    , keywords =

    Comparisons of the Emission-Line and Continuum Properties of Broad Absorption Line and Normal Quasi-stellar Objects. , keywords =. doi:10.1086/170020 , adsurl =

    doi

  232. [242]

    , keywords =

    Reionization history constraints from neural network based predictions of high-redshift quasar continua. , keywords =. doi:10.1093/mnras/staa505 , archivePrefix =. 1912.01050 , primaryClass =

    1912 arXiv

  233. [243]

    , keywords =

    Quantitative Constraints on the Reionization History from the IGM Damping Wing Signature in Two Quasars at z > 7. , keywords =. doi:10.3847/1538-4357/aad6dc , archivePrefix =. 1802.06066 , primaryClass =

    arXiv

  234. [244]

    , keywords =

    A comparison of quasar emission reconstruction techniques for z 5.0 Lyman and Lyman transmission. , keywords =. doi:10.1093/mnras/stab572 , archivePrefix =. 2006.10744 , primaryClass =

    2006 arXiv

  235. [245]

    , keywords =

    The Sloan Digital Sky Survey Quasar Catalog: Sixteenth Data Release. , keywords =. doi:10.3847/1538-4365/aba623 , archivePrefix =. 2007.09001 , primaryClass =

    2007 arXiv

  236. [246]

    , keywords =

    The Sloan Digital Sky Survey Quasar Catalog: Fourteenth data release. , keywords =. doi:10.1051/0004-6361/201732445 , archivePrefix =. 1712.05029 , primaryClass =

    arXiv

  237. [247]

    , keywords =

    The Sloan Digital Sky Survey Quasar Catalog: Twelfth data release. , keywords =. doi:10.1051/0004-6361/201527999 , archivePrefix =. 1608.06483 , primaryClass =

    arXiv

  238. [248]

    , keywords =

    The Eleventh and Twelfth Data Releases of the Sloan Digital Sky Survey: Final Data from SDSS-III. , keywords =. doi:10.1088/0067-0049/219/1/12 , archivePrefix =. 1501.00963 , primaryClass =

    arXiv

  239. [249]

    arXiv e-prints , keywords =

    Blind QSO reconstruction challenge: Exploring methods to reconstruct the Ly emission line of QSOs. arXiv e-prints , keywords =. doi:10.48550/arXiv.2404.01556 , archivePrefix =. 2404.01556 , primaryClass =

    doi arXiv

  240. [250]

    arXiv e-prints , keywords =

    LongNet: Scaling Transformers to 1,000,000,000 Tokens. arXiv e-prints , keywords =. doi:10.48550/arXiv.2307.02486 , archivePrefix =. 2307.02486 , primaryClass =

    doi arXiv

  241. [251]

    arXiv e-prints , keywords =

    FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness. arXiv e-prints , keywords =. doi:10.48550/arXiv.2205.14135 , archivePrefix =. 2205.14135 , primaryClass =

    doi arXiv

  242. [252]

    arXiv e-prints , keywords =

    Data Movement Is All You Need: A Case Study on Optimizing Transformers. arXiv e-prints , keywords =. doi:10.48550/arXiv.2007.00072 , archivePrefix =. 2007.00072 , primaryClass =

    2007 doi arXiv

  243. [253]

    , keywords =

    Classification and analysis of emission-line galaxies using mean field independent component analysis. , keywords =. doi:10.1093/mnras/stt151 , archivePrefix =. 1301.5930 , primaryClass =

    arXiv

  244. [254]

    , keywords =

    BAL and non-BAL quasars: continuum, emission, and absorption properties establish a common parent sample. , keywords =. doi:10.1093/mnras/staa130 , archivePrefix =. 1912.08700 , primaryClass =

    1912 arXiv

  245. [255]

    arXiv e-prints , keywords =

    Attention Is All You Need. arXiv e-prints , keywords =. doi:10.48550/arXiv.1706.03762 , archivePrefix =. 1706.03762 , primaryClass =

    doi arXiv

  246. [256]

    arXiv e-prints , keywords =

    Recurrent Models of Visual Attention. arXiv e-prints , keywords =. doi:10.48550/arXiv.1406.6247 , archivePrefix =. 1406.6247 , primaryClass =

    doi arXiv

  247. [257]

    , keywords =

    Towards an astronomical foundation model for stars with a transformer-based model. , keywords =. doi:10.1093/mnras/stad3015 , archivePrefix =. 2308.10944 , primaryClass =

    arXiv

  248. [258]

    Supermassive Black Holes with High Accretion Rates in Active Galactic Nuclei. X. Optical Variability Characteristics. , keywords =. doi:10.3847/1538-4357/ab16e8 , archivePrefix =. 1904.03393 , primaryClass =

    1904 arXiv

  249. [259]

    , keywords =

    The Sloan Digital Sky Survey Reverberation Mapping Project: Sample Characterization. , keywords =. doi:10.3847/1538-4365/ab074f , archivePrefix =. 1810.01447 , primaryClass =

    arXiv

  250. [260]

    , keywords =

    Observational window effects on multi-object reverberation mapping. , keywords =. doi:10.1093/mnras/stac2263 , archivePrefix =. 2203.04518 , primaryClass =

    arXiv

  251. [261]

    , keywords =

    The Sloan Digital Sky Survey Reverberation Mapping Project: Key Results. , keywords =. doi:10.3847/1538-4365/ad3936 , archivePrefix =. 2305.01014 , primaryClass =

    arXiv

  252. [262]

    , keywords =

    The Relationship between Luminosity and Broad-Line Region Size in Active Galactic Nuclei. , keywords =. doi:10.1086/431275 , archivePrefix =. astro-ph/0504484 , primaryClass =

    arXiv

  253. [263]

    , keywords =

    Black Hole Mass and Growth Rate at High Redshift. , keywords =. doi:10.1086/523035 , archivePrefix =. 0708.3787 , primaryClass =

    arXiv

  254. [264]

    , keywords =

    Reverberation Measurements for 17 Quasars and the Size-Mass-Luminosity Relations in Active Galactic Nuclei. , keywords =. doi:10.1086/308704 , archivePrefix =. astro-ph/9911476 , primaryClass =

    arXiv

  255. [265]

    , keywords =

    Biases in Virial Black Hole Masses: An SDSS Perspective. , keywords =. doi:10.1086/587475 , archivePrefix =. 0709.3098 , primaryClass =

    arXiv

  256. [266]

    , keywords =

    A Catalog of Quasar Properties from Sloan Digital Sky Survey Data Release 7. , keywords =. doi:10.1088/0067-0049/194/2/45 , archivePrefix =. 1006.5178 , primaryClass =

    arXiv

  257. [267]

    ascl:1809.008 , adsurl =

    PyQSOFit: Python code to fit the spectrum of quasars. ascl:1809.008 , adsurl =

  258. [268]

    , keywords =

    A Catalog of Quasar Properties from Sloan Digital Sky Survey Data Release 16. , keywords =. doi:10.3847/1538-4365/ac9ead , archivePrefix =. 2209.03987 , primaryClass =

    arXiv

  259. [269]

    , keywords =

    Black Hole Mass and Eddington Ratio as Drivers for the Observable Properties of Radio-loud and Radio-quiet QSOs. , keywords =. doi:10.1086/324486 , archivePrefix =. astro-ph/0109317 , primaryClass =

    arXiv

  260. [270]

    A transformer-based embedding for the representation of light curves

    ASTROMER. A transformer-based embedding for the representation of light curves. , keywords =. doi:10.1051/0004-6361/202243928 , archivePrefix =. 2205.01677 , primaryClass =

    arXiv

  261. [271]

    , keywords =

    The diversity of quasars unified by accretion and orientation. , keywords =. doi:10.1038/nature13712 , archivePrefix =. 1409.2887 , primaryClass =

    arXiv

  262. [272]

    , keywords =

    C IV 1549 as an Eigenvector 1 Parameter for Active Galactic Nuclei. , keywords =. doi:10.1086/519916 , archivePrefix =. 0705.1895 , primaryClass =

    arXiv

  263. [273]

    , keywords =

    Average Quasar Spectra in the Context of Eigenvector 1. , keywords =. doi:10.1086/339594 , archivePrefix =. astro-ph/0201362 , primaryClass =

    arXiv

  264. [274]

    , keywords =

    Eigenvector 1: An Optimal Correlation Space for Active Galactic Nuclei. , keywords =. doi:10.1086/312717 , archivePrefix =. astro-ph/0005177 , primaryClass =

    arXiv

  265. [275]

    , keywords =

    The Emission-Line Properties of Low-Redshift Quasi-stellar Objects. , keywords =. doi:10.1086/191661 , adsurl =

    doi

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.