Pith. sign in

REVIEW 3 major objections 6 minor 75 references

A PCA-Transformer spectrum classifier finds two previously unknown optical-UV tidal disruption events in SDSS DR7, one from before February 2002.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.5

2026-07-14 04:53 UTC pith:NOPKU7PA

load-bearing objection Two real early TDEs dug out of SDSS DR7 by a PCA-Transformer spectrum classifier; the discoveries hold up on public light curves even though the training library is tiny. the 3 major comments →

arxiv 2607.11539 v1 pith:NOPKU7PA submitted 2026-07-13 astro-ph.HE astro-ph.IM

Two Earliest Optical-UV Tidal Disruption Events Hidden in the SDSS DR7 Catalog Unveiled by the Transformer-Based Spectrum Classifier

Ranfang Zheng , Zheyu Lin , Xu Kong , Ran Song , Dezheng Meng , Lulu Fan This is my paper
classification astro-ph.HE astro-ph.IM
keywords Tidal disruption eventsTransformerTransientSpectrum classificationSDSSPCAoptical-UV TDE
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

Large spectroscopic surveys contain millions of galaxy spectra, so some should have been taken while a rare tidal disruption event was flaring. The authors build a PCA-enhanced Transformer classifier that scores spectra for TDE-like blue continua and broad H/He/Bowen lines, then apply multi-band light-curve cuts to reject AGN and star-formation impostors. In the widely used SDSS DR7 they recover three candidates: one already-noted likely TDE and two new ones. One shows a UV transient already bright at the spectrum epoch (MJD 52316), making its start earlier than any previously known optical-UV TDE; the other was caught mid-outburst in CRTS with a classic TDE-H+He spectrum. The result shows that machine-learning spectrum classifiers can mine archival catalogs for events that time-domain surveys never flagged.

Core claim

A PCA-Transformer TDE spectrum classifier applied to SDSS DR7 recovers two previously unreported optical-UV TDEs (plus one known likely TDE). SDSS J124225.39+642919.0 exhibits a blue continuum and broad lines together with a GALEX UV transient present at the spectrum epoch, so the flare began before MJD 52316 and is the earliest known optical-UV TDE. SDSS J152459.70+045423.1 shows a full TDE-H+He spectrum taken during a CRTS optical outburst whose onset lies between MJD 54269 and 54476.

What carries the argument

PCA-Transformer (MgFormer) spectrum classifier: spectra are projected onto fixed galaxy/transient/stellar PCA templates to form dual-channel coefficient matrices that a multi-group Transformer then scores for TDE probability; high-score candidates are further vetted by burst-like multi-band light curves and absence of persistent AGN variability.

Load-bearing premise

The small library of real TDE templates and the chosen host-mixing plus light-curve cuts are assumed to be enough to guarantee that the two retained objects are genuine TDEs rather than rare AGN or star-formation impostors that pass the same empirical filters.

What would settle it

Deep multi-epoch UV or soft X-ray imaging of the two hosts that shows either no fading continuum after 2002/2008 or persistent stochastic variability and strong narrow [O III] that would reclassify them as AGN.

Watch this falsifier — get emailed when new claim-graph text bears on it.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. The paper presents a PCA-enhanced Transformer (MgFormer) classifier for identifying TDE-like optical spectra, trained on WISeREP/NGSF transient templates mixed with SDSS host galaxies and evaluated on a held-out synthetic set (precision 0.88, recall 0.99 for TDEs). Applied to ~3×10^5 quality-selected SDSS DR7 galaxy spectra, the model yields 14 candidates with TDE score >0.5; multi-wavelength vetting (GALEX, CRTS, ZTF, WISE) plus emission-line diagnostics retain two newly reported TDEs—SDSS J124225.39+642919.0 (UV transient co-temporal with the spectrum at MJD 52316, claimed as the earliest optical-UV TDE) and SDSS J152459.70+045423.1 (TDE-H+He spectrum during a CRTS outburst with start 54269 < MJD < 54476)—plus one previously reported likely TDE. The pipeline is released open-source.

Significance. If the two new events hold under independent scrutiny, the work is a clear contribution: it recovers serendipitous TDE spectra from a widely used archival catalog, pushes the earliest known optical-UV TDE occurrence to before 2002 February, and demonstrates a practical ML route for spectroscopic TDE searches ahead of DESI-scale samples. Strengths include a clean train/test split with source-level separation for transients, explicit PCA templates and mixing protocol, open code, and—most importantly—authentication of the two discoveries by public multi-band light curves and spectral fits that are independent of the training labels. The method paper and the discovery paper are both of interest to the TDE and time-domain communities.

major comments (3)
  1. [§3.1, Table 6, Fig. 3] §3.1 and Table 6 vs §2.4/Fig. 3: On the synthetic test set the TDE precision is 0.88, but among the 14 real SDSS candidates with score >0.5 only 2–3 survive multi-wavelength vetting (real-world precision ~0.14–0.21). This domain-shift gap is load-bearing for the claim that the classifier is a powerful discovery tool for large catalogs. The main text should quantify and discuss it explicitly (expected false-positive rate under the adopted threshold, role of AGN/SF impostors, and how purity would scale to DESI), rather than leaving the discrepancy largely to Appendix B.
  2. [§2.2.1, Table 1, Appendix B] §2.2.1 / Table 1: The TDE training library comprises only 10 unique sources (3 H, 5 H+He, 2 He; 47 spectra). Appendix B correctly notes that this scarcity, and the absence of featureless-TDE templates, limits generalization and score interpretability. Because the abstract and introduction present the classifier as a general selection method, the main text (not only the appendix) should state this limitation up front and clarify which spectroscopic subclasses the model is expected to recover or miss.
  3. [§3.1.2, Table 7] §3.1.2 and Table 7: Final promotion from candidate to TDE rests on hand-crafted photometric cuts (monotonic burst in −0.5/+2 yr, DRW τ_3σ-low thresholds, WISE χ²/d.o.f. < 10, W1−W2 < 0.8). These cuts are reasonable and the two retained objects pass cleanly, but the paper should test or at least discuss sensitivity of the final sample to the window and threshold choices, and state that authentication—not the Transformer score—is what establishes the TDE nature. A short robustness paragraph would strengthen the discovery claims without changing the conclusions.
minor comments (6)
  1. [§2.3.3, Eq. (1)] Eq. (1) and §2.3.3: Both transient and galaxy spectra are min–max normalized before mixing, so Scale is a ratio of normalized fluxes, not a physical continuum flux ratio. A one-sentence clarification would avoid misinterpretation when others reuse the pipeline.
  2. [§3.2] §3.2 (J1242): The He II/Hβ complex is described as hard to resolve, yet FWHMs with large uncertainties are reported. Consider stating more clearly which lines are robust detections versus tentative, and whether the object would still be classified as TDE-H (or featureless-like) under the van Velzen/Hammerstein scheme if He II is not required.
  3. [Fig. 5] Figure 5 / light-curve panels: Magnitudes are offset for display; the offset values are given in the legend but a brief note in the caption that fluxes are not host-subtracted (already in the text) would help casual readers.
  4. [Table 7] Table 7 is very dense; a short legend defining each column’s decision rule (already partly in the notes) or moving the full decision tree to the appendix would improve readability.
  5. [Tables 6–7] Minor consistency: SDSS names sometimes differ by 0.01 s or last digit between Table 6, Table 7, and the text (e.g., J074820.66 vs J074820.67). Unify identifiers.
  6. [Abstract] The abstract’s phrase “inspiring discoveries” is informal for a journal abstract; a more neutral wording would match the rest of the paper’s tone.

Circularity Check

0 steps flagged

No significant circularity: classifier trained on external WISeREP templates + synthetic mixes; final TDE claims rest on independent GALEX/CRTS photometry and standard empirical spectral criteria.

full rationale

The derivation chain is self-contained and non-circular. Training labels and PCA templates come from external WISeREP/NGSF TDE spectra (Table 1: only 10 unique TDE sources) mixed with SDSS galaxies/stars under explicit scales (Eq. 1, Table 3); evaluation precision/recall (0.88/0.99) is measured on held-out synthetic mixes never used for the real-data claims. Application yields 14 candidates via TDE score >0.5; the two confirmed TDEs (J1242, J1524) are authenticated solely by post-hoc multi-wavelength diagnostics (blue continuum + broad H/He/Bowen lines without strong [O III]; GALEX UV transient co-temporal with MJD 52316 that fades; CRTS optical outburst bracketing the spectrum epoch; DRW τ and WISE χ²/W1–W2 cuts ruling out stochastic AGN) that match literature criteria of van Velzen et al. (2021) and Hammerstein et al. (2023). These photometry and line diagnostics were never inputs to the model. The sole self-citation (prior photometric TTC of Zheng et al. 2026) is motivational only and not load-bearing for either the spectrum architecture (MgFormer from Wen et al. 2024) or the discoveries. No equation, fit, or uniqueness claim reduces a prediction to its own inputs by construction. Template scarcity is a purity limitation, not circularity.

Axiom & Free-Parameter Ledger

6 free parameters · 4 axioms · 0 invented entities

The central discoveries rest on a modest set of free modeling choices (PCA ranks, mixing ranges, score threshold, light-curve cuts) and standard domain assumptions about TDE spectral classes and AGN variability diagnostics. No new physical entities are postulated; the classifier is a computational tool, not an invented particle or force.

free parameters (6)
  • number of galaxy PCA components = 86
    Fixed at 86 to match the meansk86 taxonomy; choice affects the dual-channel feature matrix fed to the Transformer.
  • number of transient/stellar PCA components = 20
    Retained at 20 (~98 % variance); directly sets the second channel of the (2,106) input matrix.
  • TDE–host mixing-scale ranges = 0.0–2.0 (TDE)
    Hand-chosen intervals (0–1.5/2.0 for TDEs, up to 0–10 for SNe) control the synthetic training distribution (Table 3).
  • TDE score selection threshold = 0.5
    Candidates kept only if score > 0.5; arbitrary but conventional half-probability cut that yields the 14 objects later vetted.
  • light-curve burst window and DRW/χ² cuts = –0.5/+2 yr, χ²>10, W1–W2>0.8
    Monotonic rise/fall within –0.5/+2 yr, τ_DRW,3σ-low > bin size, WISE χ²/d.o.f. > 10, W1–W2 > 0.8 used to accept/reject candidates (Section 3.1.2).
  • MgFormer hyper-parameters (multi_group, emb_size, nlayers, etc.) = multi_group=[1,2], emb_size=128, nlayers=2
    Default values taken from Wen et al. (2024) and listed in Table 5; control model capacity and training dynamics.
axioms (4)
  • domain assumption Optical TDE spectra are adequately described by the four empirical classes (TDE-H, TDE-He, TDE-H+He, featureless) defined by the presence/absence of broad Balmer, He and Bowen lines plus a blue continuum.
    Adopted from van Velzen et al. (2021) and Hammerstein et al. (2023); used both for training labels and for final spectral classification of the two new objects.
  • domain assumption A single monotonic optical/UV/MIR burst near the spectrum epoch, combined with the absence of long-term stochastic DRW variability and W1–W2 < 0.8, is sufficient to exclude AGN impostors.
    Explicitly applied in Section 3.1.2 to promote two candidates to confirmed TDEs and demote six others to AGNs.
  • ad hoc to paper Linear superposition of rest-frame, min-max-normalized transient and host spectra with additive Gaussian noise scaled by (1+Scale)⁻¹ realistically approximates observed SDSS spectra of ongoing TDEs.
    Core of the training-set construction (Eq. 1 and surrounding text); never validated against real simultaneous host+TDE spectra.
  • domain assumption The meansk86 unsupervised taxonomy of SDSS galaxies supplies a complete, non-leaking basis for host PCA templates.
    Used for the final re-training stage (Section 2.5.1); assumed free of test-set leakage after the authors switch from training-only PCs.

pith-pipeline@v1.1.0-grok45 · 28959 in / 3430 out tokens · 40037 ms · 2026-07-14T04:53:01.842822+00:00 · methodology

0 comments
read the original abstract

Optical spectroscopic features are decisive in the current identification of optical-UV tidal disruption events (TDEs). Regarding that the TDE estimated occurrence rate is $10^{-5}-10^{-4}$ galaxy$^{-1}$ yr$^{-1}$ by both theoretical and observational methods, large optical spectroscopic catalogs with >10$^5$ galaxy spectra can include some serendipitous spectra with TDE spectroscopic features, which can be found after building a useful selection method. We hereby introduce a principal component analysis enhanced Transformer TDE spectrum classifier which achieves a precision of 0.88 and a recall of 0.99 on our evaluation dataset, and report its inspiring discoveries in the widely-used SDSS DR7 catalog: two newly discovered TDEs and one reported likely TDE. For SDSS J124225.39+642919.0, we confirm the presence of a UV transient in GALEX catalog when the spectrum was taken, and its occurrence time should be earlier than the spectrum observation time, MJD < 52316 (February 11, 2002), making it the earliest optical-UV TDE discovered by now. For SDSS J152459.70+045423.1, its spectrum matches all features of the TDE-H+He spectrum, and was taken during an optical outburst recorded by the Catalina Real-time Transient Survey. The start of this outburst lies in 54269 < MJD < 54476 (June 18, 2007 - January 11, 2008), making it one of the earliest among the reported optical TDEs. The discovery of two new TDEs highlights the power of machine-learning based classifiers in digging out buried treasures in large-volume catalogs, and marks a new method for discovering optical-UV TDEs.

Figures

Figures reproduced from arXiv: 2607.11539 by Dezheng Meng, Lulu Fan, Ranfang Zheng, Ran Song, Xu Kong, Zheyu Lin.

Figure 1
Figure 1. Figure 1: Schematic illustration of the preparation procedure for the training and test datasets. We first select spectra of the transient source and its host galaxy. After applying redshift correction, each spectrum is normalized by its maximum flux. The two spectra are then superimposed using different mixing scales, where the mixing scale represents the flux ratio of the host galaxy to the transient source during… view at source ↗
Figure 2
Figure 2. Figure 2: Cumulative explained variance of the PCA for transient source spectra. Retaining 20 principal components captures approximately 98% of the total variance, indicating that the dominant spectral features are effectively preserved. balancing noise restoration with the preservation of intrinsic spectral features. After this step, wavelet-based denoising is applied to the noise-augmented mixed spectra using the… view at source ↗
Figure 3
Figure 3. Figure 3: The confusion matrices that show the capability of distinguishing TDEs from non-TDEs. For the independent test set used for performance evaluation, the validation model achieved a precision of 0.88 and a recall of 0.99 for TDE classification. 0.0 0.2 0.4 0.6 0.8 1.0 TDE Score 10 0 10 1 10 2 10 3 10 4 10 5 Number of Sources Total Sources: 298195 [0.0, 0.5]: 298181 (0.5, 1.0]: 14 TDE Score Distribution [PIT… view at source ↗
Figure 4
Figure 4. Figure 4: TDE Score distribution across all evaluated sources. The vast majority of the population is concentrated at low confidence levels (< 0.1). The blue dotted line indicates the selection threshold (TDE Score = 0.5), above which sources are identified as potential candidates. Only 14 of the total samples have TDE score>0.5. firm they were taken on the spirals of nearby galaxies. The blue continuum (fλ ∝ λ α an… view at source ↗
Figure 5
Figure 5. Figure 5: Top: UV (GALEX) and optical (CRTS) light curves of two newly discovered TDEs around the SDSS spectrum observation time (dotted black line). Magnitudes are corrected for Galactic extinction but not host-subtracted, and on which offsets are applied for clearer display. Bottom: The GALEX FUV and NUV “intensity” map of SDSS J124225.39+642919 in three epochs, which is created by dividing the count map by the re… view at source ↗
Figure 6
Figure 6. Figure 6: The rest-frame spectra of SDSS J124225.39+642919.0, SDSS J152459.70+045423.1 and SDSS J074820.67+471214.2. In particular, SDSS J152459.70+045423.1 and SDSS J124225.39+642919.0 are first reported in this article and spectral fitting is performed on them. The best-fit continuum is plotted as a gray dashed line. Shaded regions indicate individual emission-line components: broad Hα (purple) and He I (orange), … view at source ↗

discussion (0)

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

Reference graph

Works this paper leans on

75 extracted references · 10 canonical work pages · 8 internal anchors

  1. [4]

    1996, title Detection of an extremely soft X-ray outburst in the HII-like nucleus of NGC 5905

    Bade , N., Komossa , S., & Dahlem , M. 1996, title Detection of an extremely soft X-ray outburst in the HII-like nucleus of NGC 5905. , , 309, L35

    1996

  2. [10]

    2015, title PCANet: A Simple Deep Learning Baseline for Image Classification? IEEE Transactions on Image Processing, 24, 5017, 10.1109/TIP.2015.2475625

    Chan, T.-H., Jia, K., Gao, S., et al. 2015, title PCANet: A Simple Deep Learning Baseline for Image Classification? IEEE Transactions on Image Processing, 24, 5017, 10.1109/TIP.2015.2475625

    doi:10.1109/tip.2015.2475625 2015

  3. [11]

    G., et al

    DESI Collaboration , Abdul-Karim, M., Adame, A. G., et al. 2025, Data Release 1 of the Dark Energy Spectroscopic Instrument, 2503.14745

    Pith/arXiv arXiv 2025

  4. [23]

    2022, title The Next Generation SuperFit (NGSF) tool is now available for online execution on WISeREP , Transient Name Server AstroNote, 191, 1

    Goldwasser , S., Yaron , O., Sass , A., et al. 2022, title The Next Generation SuperFit (NGSF) tool is now available for online execution on WISeREP , Transient Name Server AstroNote, 191, 1

    2022

  5. [32]

    2014, title Adam: A Method for Stochastic Optimization, International Conference on Learning Representations

    Kingma, D., & Ba, J. 2014, title Adam: A Method for Stochastic Optimization, International Conference on Learning Representations

    2014

  6. [34]

    2023, title Operator learning with PCA-Net: upper and lower complexity bounds, Journal of Machine Learning Research, 24, 1

    Lanthaler, S. 2023, title Operator learning with PCA-Net: upper and lower complexity bounds, Journal of Machine Learning Research, 24, 1. http://jmlr.org/papers/v24/23-0478.html

    2023

  7. [44]

    E., & Ferland , G

    Osterbrock , D. E., & Ferland , G. J. 2006, Astrophysics of gaseous nebulae and active galactic nuclei (University Science Books)

    2006

  8. [45]

    2019, title PyTorch: An Imperative Style, High-Performance Deep Learning Library, in Advances in Neural Information Processing Systems 32, ed

    Paszke, A., Gross, S., Massa, F., et al. 2019, title PyTorch: An Imperative Style, High-Performance Deep Learning Library, in Advances in Neural Information Processing Systems 32, ed. H. Wallach, H. Larochelle, A. Beygelzimer, F. d Alch\' e -Buc, E. Fox, & R. Garnett (Curran Associates, Inc.), 8024--8035. http://papers.neurips.cc/paper/9015-pytorch-an \ -...

    2019

  9. [61]

    Wen, J., Zhang, N., Lu, X., Hu, Z., & Huang, H. 2024, title Mgformer: Multi-group transformer for multivariate time series classification, Engineering Applications of Artificial Intelligence, 133, 108633, https://doi.org/10.1016/j.engappai.2024.108633

    work page doi:10.1016/j.engappai.2024.108633 2024

  10. [62]

    L., Eisenhardt , P

    Wright , E. L., Eisenhardt , P. R. M., Mainzer , A. K., et al. 2010, title The Wide-field Infrared Survey Explorer (WISE): Mission Description and Initial On-orbit Performance , , 140, 1868. 1008.0031

    Pith/arXiv arXiv 2010

  11. [67]

    Insights from the "Red devil" AT 2022fpx: A Dust-reddened Family of Tidal Disruption Events Excluded by Their Apparent Red Color?

    Insights from the ``Red Devil'' AT 2022fpx: A Dust-reddened Family of Tidal Disruption Events Excluded by Their Apparent Red Color?. , keywords =. doi:10.3847/1538-4357/adef10 , archivePrefix =. 2507.04834 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.3847/1538-4357/adef10

  12. [68]

    , keywords =

    Long Fading Mid-infrared Emission in Transient Coronal Line Emitters: Dust Echo of a Tidal Disruption Flare. , keywords =. doi:10.3847/0004-637X/832/2/188 , archivePrefix =. 1605.05145 , primaryClass =

    Pith/arXiv arXiv doi:10.3847/0004-637x/832/2/188

  13. [69]

    , keywords =

    The WISE Detection of an Infrared Echo in Tidal Disruption Event ASASSN-14li. , keywords =. doi:10.3847/2041-8205/828/1/L14 , archivePrefix =. 1605.04640 , primaryClass =

    Pith/arXiv arXiv doi:10.3847/2041-8205/828/1/l14 2041

  14. [70]

    , keywords =

    Discovery of Transient Infrared Emission from Dust Heated by Stellar Tidal Disruption Flares. , keywords =. doi:10.3847/0004-637X/829/1/19 , archivePrefix =. 1605.04304 , primaryClass =

    Pith/arXiv arXiv doi:10.3847/0004-637x/829/1/19

  15. [71]

    , keywords =

    The Zwicky Transient Facility: Surveys and Scheduler. , keywords =. doi:10.1088/1538-3873/ab0c2a , archivePrefix =. 1905.02209 , primaryClass =

    Pith/arXiv arXiv doi:10.1088/1538-3873/ab0c2a 1905

  16. [72]

    , keywords =

    Are the Variations in Quasar Optical Flux Driven by Thermal Fluctuations?. , keywords =. doi:10.1088/0004-637X/698/1/895 , archivePrefix =. 0903.5315 , primaryClass =

    Pith/arXiv arXiv doi:10.1088/0004-637x/698/1/895

  17. [73]

    , keywords =

    Modeling the Time Variability of SDSS Stripe 82 Quasars as a Damped Random Walk. , keywords =. doi:10.1088/0004-637X/721/2/1014 , archivePrefix =. 1004.0276 , primaryClass =

    Pith/arXiv arXiv doi:10.1088/0004-637x/721/2/1014

  18. [74]

    Mid-infrared Selection of Active Galactic Nuclei with the Wide-Field Infrared Survey Explorer. I. Characterizing WISE-selected Active Galactic Nuclei in COSMOS. , keywords =. doi:10.1088/0004-637X/753/1/30 , archivePrefix =. 1205.0811 , primaryClass =

    Pith/arXiv arXiv doi:10.1088/0004-637x/753/1/30

  19. [75]

    , keywords =

    The Seventh Data Release of the Sloan Digital Sky Survey. , keywords =. doi:10.1088/0067-0049/182/2/543 , archivePrefix =. 0812.0649 , primaryClass =

    Pith/arXiv arXiv doi:10.1088/0067-0049/182/2/543

  20. [76]

    , keywords =

    A Continuum of H- to He-rich Tidal Disruption Candidates With a Preference for E+A Galaxies. , keywords =. doi:10.1088/0004-637X/793/1/38 , archivePrefix =. 1405.1415 , primaryClass =

    Pith/arXiv arXiv doi:10.1088/0004-637x/793/1/38

  21. [77]

    , keywords =

    Detection of an extremely soft X-ray outburst in the HII-like nucleus of NGC 5905. , keywords =

  22. [78]

    iPTF16fnl: a faint and fast tidal disruption event in an E+A galaxy

    iPTF16fnl: A Faint and Fast Tidal Disruption Event in an E+A Galaxy. , keywords =. doi:10.3847/1538-4357/aa7579 , archivePrefix =. 1703.00965 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.3847/1538-4357/aa7579

  23. [79]

    2025 , eprint=

    Data Release 1 of the Dark Energy Spectroscopic Instrument , author=. 2025 , eprint=

    2025

  24. [80]

    Large-Amplitude X-ray Outbursts from Galactic Nuclei: A Systematic Survey Using ROSAT Archival Data

    Large-Amplitude X-Ray Outbursts from Galactic Nuclei: A Systematic Survey using ROSAT Archival Data. , keywords =. doi:10.1086/342280 , archivePrefix =. astro-ph/0206291 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1086/342280

  25. [81]

    , keywords =

    First Results from the Catalina Real-Time Transient Survey. , keywords =. doi:10.1088/0004-637X/696/1/870 , archivePrefix =. 0809.1394 , primaryClass =

    Pith/arXiv arXiv doi:10.1088/0004-637x/696/1/870

  26. [82]

    , keywords =

    Evolution of tidal disruption candidates discovered by XMM-Newton. , keywords =. doi:10.1051/0004-6361:200810110 , archivePrefix =. 0807.4452 , primaryClass =

    Pith/arXiv arXiv doi:10.1051/0004-6361:200810110

  27. [83]

    , keywords =

    Correcting for the Effects of Interstellar Extinction. , keywords =. doi:10.1086/316293 , archivePrefix =. astro-ph/9809387 , primaryClass =

    Pith/arXiv arXiv doi:10.1086/316293

  28. [84]

    , keywords =

    An ultraviolet-optical flare from the tidal disruption of a helium-rich stellar core. , keywords =. doi:10.1038/nature10990 , archivePrefix =. 1205.0252 , primaryClass =

    Pith/arXiv arXiv doi:10.1038/nature10990

  29. [85]

    , keywords =

    UV/Optical Detections of Candidate Tidal Disruption Events by GALEX and CFHTLS. , keywords =. doi:10.1086/529008 , archivePrefix =. 0712.4149 , primaryClass =

    Pith/arXiv arXiv doi:10.1086/529008

  30. [86]

    , keywords =

    Ultraviolet Detection of the Tidal Disruption of a Star by a Supermassive Black Hole. , keywords =. doi:10.1086/509918 , archivePrefix =. astro-ph/0612069 , primaryClass =

    Pith/arXiv arXiv doi:10.1086/509918

  31. [87]

    , keywords =

    Luminous Thermal Flares from Quiescent Supermassive Black Holes. , keywords =. doi:10.1088/0004-637X/698/2/1367 , archivePrefix =. 0904.1596 , primaryClass =

    Pith/arXiv arXiv doi:10.1088/0004-637x/698/2/1367

  32. [88]

    , keywords =

    The Final Season Reimagined: 30 Tidal Disruption Events from the ZTF-I Survey. , keywords =. doi:10.3847/1538-4357/aca283 , archivePrefix =. 2203.01461 , primaryClass =

    Pith/arXiv arXiv doi:10.3847/1538-4357/aca283

  33. [89]

    , keywords =

    Possible power source of Seyfert galaxies and QSOs. , keywords =. doi:10.1038/254295a0 , adsurl =

    doi:10.1038/254295a0

  34. [90]

    , keywords =

    ASASSN-14ae: a tidal disruption event at 200 Mpc. , keywords =. doi:10.1093/mnras/stu1922 , archivePrefix =. 1405.1417 , primaryClass =

    Pith/arXiv arXiv doi:10.1093/mnras/stu1922

  35. [91]

    , keywords =

    Six months of multiwavelength follow-up of the tidal disruption candidate ASASSN-14li and implied TDE rates from ASAS-SN. , keywords =. doi:10.1093/mnras/stv2486 , archivePrefix =. 1507.01598 , primaryClass =

    Pith/arXiv arXiv doi:10.1093/mnras/stv2486

  36. [92]

    , keywords =

    The giant X-ray outbursts in NGC 5905 and IC 3599:() hfill Follow-up observations and outburst scenarios. , keywords =. doi:10.48550/arXiv.astro-ph/9901141 , archivePrefix =. astro-ph/9901141 , primaryClass =

    Pith/arXiv arXiv doi:10.48550/arxiv.astro-ph/9901141

  37. [93]

    , keywords =

    The distribution and consumption rate of stars around a massive, collapsed object. , keywords =. doi:10.1086/154925 , adsurl =

    doi:10.1086/154925

  38. [94]

    doi:10.5281/zenodo.7472061 , version =

    MillionConcepts/gPhoton2: v3.0.0a0. doi:10.5281/zenodo.7472061 , version =

    work page doi:10.5281/zenodo.7472061

  39. [95]

    , keywords =

    Rates of tidal disruption of stars by massive central black holes. , keywords =. doi:10.1046/j.1365-8711.1999.02853.x , archivePrefix =. astro-ph/9902032 , primaryClass =

    doi:10.1046/j.1365-8711.1999.02853.x 1999

  40. [96]

    , keywords =

    The Galaxy Evolution Explorer: A Space Ultraviolet Survey Mission. , keywords =. doi:10.1086/426387 , archivePrefix =. astro-ph/0411302 , primaryClass =

    Pith/arXiv arXiv doi:10.1086/426387

  41. [97]

    , year = 1974, month = feb, volume =

    Absolute Spectral Energy Distributions for White Dwarfs. , year = 1974, month = feb, volume =. doi:10.1086/190287 , adsurl =

    doi:10.1086/190287 1974

  42. [98]

    Astrophysics of gaseous nebulae and active galactic nuclei

  43. [99]

    Planck intermediate results. XLVIII. Disentangling Galactic dust emission and cosmic infrared background anisotropies. , keywords =. doi:10.1051/0004-6361/201629022 , archivePrefix =. 1605.09387 , primaryClass =

    Pith/arXiv arXiv doi:10.1051/0004-6361/201629022

  44. [100]

    , keywords =

    Tidal disruption of stars by black holes of 10 ^ 6 - 10 ^ 8 solar masses in nearby galaxies. , keywords =. doi:10.1038/333523a0 , adsurl =

    doi:10.1038/333523a0

  45. [101]

    , keywords =

    Rates of stellar tidal disruption as probes of the supermassive black hole mass function. , keywords =. doi:10.1093/mnras/stv2281 , archivePrefix =. 1410.7772 , primaryClass =

    Pith/arXiv arXiv doi:10.1093/mnras/stv2281

  46. [102]

    , keywords =

    Tidal disruption rates of stars in observed galaxies. , keywords =. doi:10.1046/j.1365-8711.1999.02445.x , archivePrefix =. astro-ph/9812389 , primaryClass =

    doi:10.1046/j.1365-8711.1999.02445.x 1999

  47. [103]

    , keywords =

    On the Mass and Luminosity Functions of Tidal Disruption Flares: Rate Suppression due to Black Hole Event Horizons. , keywords =. doi:10.3847/1538-4357/aa998e , archivePrefix =. 1707.03458 , primaryClass =

    Pith/arXiv arXiv doi:10.3847/1538-4357/aa998e

  48. [104]

    , keywords =

    Optical Discovery of Probable Stellar Tidal Disruption Flares. , keywords =. doi:10.1088/0004-637X/741/2/73 , archivePrefix =. 1009.1627 , primaryClass =

    Pith/arXiv arXiv doi:10.1088/0004-637x/741/2/73

  49. [105]

    Measurement of the rate of stellar tidal disruption flares

    Measurement of the Rate of Stellar Tidal Disruption Flares. , keywords =. doi:10.1088/0004-637X/792/1/53 , archivePrefix =. 1407.6425 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1088/0004-637x/792/1/53

  50. [106]

    , keywords =

    Seventeen Tidal Disruption Events from the First Half of ZTF Survey Observations: Entering a New Era of Population Studies. , keywords =. doi:10.3847/1538-4357/abc258 , archivePrefix =. 2001.01409 , primaryClass =

    Pith/arXiv arXiv doi:10.3847/1538-4357/abc258 2001

  51. [107]

    , keywords =

    Revised Rates of Stellar Disruption in Galactic Nuclei. , keywords =. doi:10.1086/379767 , archivePrefix =. astro-ph/0305493 , primaryClass =

    Pith/arXiv arXiv doi:10.1086/379767

  52. [108]

    , keywords =

    Extreme Coronal Line Emitters: Tidal Disruption of Stars by Massive Black Holes in Galactic Nuclei?. , keywords =. doi:10.1088/0004-637X/749/2/115 , archivePrefix =. 1202.1064 , primaryClass =

    Pith/arXiv arXiv doi:10.1088/0004-637x/749/2/115

  53. [109]

    , keywords =

    The Wide-field Infrared Survey Explorer (WISE): Mission Description and Initial On-orbit Performance. , keywords =

  54. [110]

    , keywords =

    Tidal Disruption Event Demographics with the Zwicky Transient Facility: Volumetric Rates, Luminosity Function, and Implications for the Local Black Hole Mass Function. , keywords =. doi:10.3847/2041-8213/acf216 , archivePrefix =. 2303.06523 , primaryClass =

    Pith/arXiv arXiv doi:10.3847/2041-8213/acf216 2041

  55. [111]

    , keywords =

    WISeREP An Interactive Supernova Data Repository. , keywords =. doi:10.1086/666656 , archivePrefix =. 1204.1891 , primaryClass =

    Pith/arXiv arXiv doi:10.1086/666656

  56. [112]

    , keywords =

    TTC: Transformer-based TDE Classifier for the Wide Field Survey Telescope (WFST). , keywords =. doi:10.3847/1538-4357/ae3153 , archivePrefix =. 2512.24782 , primaryClass =

    doi:10.3847/1538-4357/ae3153

  57. [113]

    Mgformer: Multi-group transformer for multivariate time series classification , url =

    Jianfeng Wen and Nan Zhang and Xuzhe Lu and Zhongyi Hu and Hui Huang , doi =. Mgformer: Multi-group transformer for multivariate time series classification , url =. Engineering Applications of Artificial Intelligence , keywords =. 2024 , bdsk-url-1 =

    2024

  58. [114]

    Unsupervised classification of SDSS galaxy spectra

    Unsupervised classification of SDSS galaxy spectra. , keywords =. doi:10.1051/0004-6361/202040046 , archivePrefix =. 2103.05928 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1051/0004-6361/202040046

  59. [115]

    , keywords =

    PS16dtm: A Tidal Disruption Event in a Narrow-line Seyfert 1 Galaxy. , keywords =. doi:10.3847/1538-4357/aa77f7 , archivePrefix =. 1703.07816 , primaryClass =

    Pith/arXiv arXiv doi:10.3847/1538-4357/aa77f7

  60. [116]

    , keywords =

    Infrared Echo and Late-stage Rebrightening of Nuclear Transient Ps1-10adi: Exploring the Torus with Tidal Disruption Events in Active Galactic Nuclei. , keywords =. doi:10.3847/1538-4357/aaf6b2 , archivePrefix =. 1812.01295 , primaryClass =

    Pith/arXiv arXiv doi:10.3847/1538-4357/aaf6b2

  61. [117]

    Light curves of tidal disruption events in active galactic nuclei

    Light Curves of Tidal Disruption Events in Active Galactic Nuclei. , keywords =. doi:10.3847/1538-4357/abb776 , archivePrefix =. 2004.06234 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.3847/1538-4357/abb776 2004

  62. [118]

    , keywords =

    Tidal Disruption Events in Active Galactic Nuclei. , keywords =. doi:10.3847/1538-4357/ab2b40 , archivePrefix =. 1904.12261 , primaryClass =

    Pith/arXiv arXiv doi:10.3847/1538-4357/ab2b40 1904

  63. [119]

    A Tidal Disruption Event Candidate Discovered in the Active Galactic Nucleus SDSS J022700.77-042020.6

    A Tidal Disruption Event Candidate Discovered in the Active Galactic Nucleus SDSS J022700.77-042020.6. , keywords =. doi:10.3847/1538-4357/ab880f , archivePrefix =. 2004.01517 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.3847/1538-4357/ab880f 2004

  64. [120]

    arXiv e-prints , keywords =

    Back from the dead: AT2019aalc as a candidate repeating TDE in an AGN. arXiv e-prints , keywords =. doi:10.48550/arXiv.2408.17419 , archivePrefix =. 2408.17419 , primaryClass =

    doi:10.48550/arxiv.2408.17419

  65. [121]

    Adam: A Method for Stochastic Optimization , journal =

    Kingma, Diederik and Ba, Jimmy , year =. Adam: A Method for Stochastic Optimization , journal =

  66. [122]

    Transient Name Server AstroNote , keywords =

    The Next Generation SuperFit (NGSF) tool is now available for online execution on WISeREP. Transient Name Server AstroNote , keywords =

  67. [123]

    PCANet: A Simple Deep Learning Baseline for Image Classification? , year=

    Chan, Tsung-Han and Jia, Kui and Gao, Shenghua and Lu, Jiwen and Zeng, Zinan and Ma, Yi , journal=. PCANet: A Simple Deep Learning Baseline for Image Classification? , year=

  68. [124]

    Journal of Machine Learning Research , year =

    Samuel Lanthaler , title =. Journal of Machine Learning Research , year =

  69. [125]

    , journal=

    Hunter, John D. , journal=. Matplotlib: A 2D Graphics Environment , year=. doi:10.1109/MCSE.2007.55 , ISSN=

    doi:10.1109/mcse.2007.55 2007

  70. [126]

    , keywords =

    Astropy: A community Python package for astronomy. , keywords =. doi:10.1051/0004-6361/201322068 , archivePrefix =. 1307.6212 , primaryClass =

    Pith/arXiv arXiv doi:10.1051/0004-6361/201322068

  71. [127]

    and Haberland, Matt and Reddy, Tyler and Cournapeau, David and Burovski, Evgeni and Peterson, Pearu and Weckesser, Warren and Bright, Jonathan and van der Walt, Stéfan J

    Virtanen, Pauli and Gommers, Ralf and Oliphant, Travis E. and Haberland, Matt and Reddy, Tyler and Cournapeau, David and Burovski, Evgeni and Peterson, Pearu and Weckesser, Warren and Bright, Jonathan and van der Walt, Stéfan J. and Brett, Matthew and Wilson, Joshua and Millman, K. Jarrod and Mayorov, Nikolay and Nelson, Andrew R. J. and Jones, Eric and K...

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

  72. [128]

    Advances in Neural Information Processing Systems 32 , editor =

    PyTorch: An Imperative Style, High-Performance Deep Learning Library , author =. Advances in Neural Information Processing Systems 32 , editor =. 2019 , publisher =

    2019

  73. [129]

    , keywords =

    SNIascore: Deep-learning Classification of Low-resolution Supernova Spectra. , keywords =. doi:10.3847/2041-8213/ac116f , archivePrefix =. 2104.12980 , primaryClass =

    Pith/arXiv arXiv doi:10.3847/2041-8213/ac116f 2041

  74. [130]

    Distinguishing Tidal Disruption Events from Impostors

    Distinguishing Tidal Disruption Events from Impostors. , keywords =. doi:10.1007/s11214-021-00829-4 , archivePrefix =. 2103.12150 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1007/s11214-021-00829-4

  75. [131]

    and Gommers, Ralf and Waselewski, Filip and Wohlfahrt, Kai and

    Lee, Gregory R. and Gommers, Ralf and Waselewski, Filip and Wohlfahrt, Kai and. PyWavelets: A Python package for wavelet analysis , journal =. 2019 , publisher =. doi:10.21105/joss.01237 , url =

    doi:10.21105/joss.01237 2019