Pith. sign in

REVIEW 2 major objections 5 minor 73 references

A composite of ~61,000 SPHEREx QSOs shows UV/optical slope −0.10 and NIR slope −1.46 that both change systematically with luminosity.

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 12:56 UTC pith:P4QXXA5M

load-bearing objection Solid large-N UV-to-NIR QSO composite from early SPHEREx; useful template with luminosity trends that hold as empirical measurements even after the host-contamination caveat is taken seriously. the 2 major comments →

arxiv 2607.10282 v1 pith:P4QXXA5M submitted 2026-07-11 astro-ph.GA

A UV-to-Near-infrared QSO Composite Spectrum from the SPHEREx All-Sky Survey

Minjin Kim , Yongjung Kim , Woong-Seob Jeong , Yujin Yang , Jeonghyun Pyo , Richard M. Feder , Bomee Lee , Dohyeong Kim
show 14 more authors
Kyuseok Oh Jong-Hak Woo Jeong Hwan Lee Yun-Ting Cheng Yi-Kuan Chiang Asantha Cooray Brendan P. Crill Olivier Dore Andreas L. Faisst Zhaoyu Huai Howard Hui Daniel C. Masters Chi H. Nguyen Michael Zemcov
This is my paper
classification astro-ph.GA
keywords quasarsactive galactic nucleicomposite spectraSPHERExnear-infraredreceding torusBaldwin effecthost-galaxy contamination
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.

This paper builds a single rest-frame spectrum spanning 0.14–4.5 µm by stacking SPHEREx spectrophotometry of roughly 61,000 type-1 SDSS quasars. The UV/optical continuum follows a power law with spectral index −0.10; the near-infrared continuum is steeper, with index −1.46, and both slopes, plus the IR-to-optical flux ratio, vary with bolometric luminosity in the sense expected if the dusty torus recedes as the central engine brightens. Broad hydrogen lines (Hα, Paβ, Paα) match Case B recombination ratios, implying little internal dust reddening, yet their equivalent widths rise with luminosity—an anti-Baldwin trend. The authors emphasize that residual host-galaxy light still shapes the long-wavelength end and must be controlled before the template is used for photo-z or selection work. The result supplies the first all-sky, UV-to-NIR empirical quasar SED that can be sliced by luminosity, black-hole mass, and Eddington ratio.

Core claim

From ~61 000 type-1 QSOs the median SPHEREx composite has UV/optical continuum α_ν = −0.10 and NIR continuum α_ν = −1.46; both indices and the IR-to-optical continuum ratio change systematically with bolometric luminosity (flatter UV/optical and steeper NIR for more luminous objects), while Hα, Paβ and Paα line ratios agree with Case B recombination and their equivalent widths increase with luminosity.

What carries the argument

The SPHEREx all-sky LVF spectrophotometric stack of ~61 000 SDSS type-1 QSOs, normalized iteratively by spectral overlap and median-combined in rest-frame bins, which yields a single high-S/N continuum and line template spanning 0.14–4.5 µm that can be further sliced by luminosity, black-hole mass and Eddington ratio.

Load-bearing premise

That cutting the optical host fraction below 10 percent at 5100 Å, plus S/N and channel-count cuts, is enough to keep residual stellar light from dominating the long-wavelength continuum shape and the luminosity trends measured there.

What would settle it

Construct identical composites after an independent, wavelength-dependent host subtraction (for example using rest-frame 1.6 µm stellar templates or higher-resolution NIR spectra) and test whether the IR slope still steepens and the IR/optical ratio still falls with rising bolometric luminosity.

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

If this is right

  • Photo-z codes for SPHEREx and similar low-resolution surveys can replace theoretical or sparsely sampled QSO SEDs with this empirical, luminosity-sliced template.
  • Color-based AGN selection can incorporate the measured variance of continuum slopes and line strengths across luminosity bins rather than a single mean SED.
  • The observed decline of IR-to-optical flux with luminosity supplies a new empirical constraint on the luminosity dependence of the torus covering factor.
  • The anti-Baldwin rise of Hα and Paschen equivalent widths with luminosity implies that low-ionization recombination lines do not soften with continuum luminosity in the same way high-ionization UV lines do.

Where Pith is reading between the lines

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

  • If residual host light is later shown to drive the NIR knee, the receding-torus interpretation of the IR/optical ratio trend will need re-examination with host-free stacks.
  • The same stacking pipeline can be re-run on later SPHEREx data releases once full-sky coverage is complete, testing whether the luminosity trends remain after the low-redshift host bias is reduced.
  • Because the composite is already sliced by Eddington ratio, it offers a ready empirical prior for models that couple accretion rate to continuum shape and line strength.

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

2 major / 5 minor

Summary. The paper constructs an empirical UV-to-NIR composite spectrum of ~61,000 type-1 SDSS QSOs (median z ≈ 1.26) from early SPHEREx spectrophotometry (0.75–5.0 µm, R ≈ 35–130), spanning rest-frame 0.14–4.5 µm. After S/N, channel-count, BAL, and f_host < 0.1 cuts, spectra are iteratively normalized and stacked (median/mean/geometric mean). The full-sample continuum is fit with α_ u,opt ≈ −0.10 (0.15 and 0.55 µm windows) and α_ u,IR ≈ −1.46 (1.2 and 2.4 µm windows). Sub-composites binned by L_bol, M_BH, and λ_Edd show that more luminous objects have flatter UV/optical and steeper NIR continua and lower IR-to-optical flux ratios, interpreted as consistent with a receding torus. Hα, He I+Paγ, Paβ, and Paα fluxes and EWs are measured; Paschen-to-Balmer ratios match Case B, while EWs increase with L_bol (anti-Baldwin). Host contamination and redshift–luminosity coupling are discussed as caveats, especially for the NIR “knee.”

Significance. This is the first large-N, all-sky UV-to-NIR QSO composite from SPHEREx, filling a long-standing gap between the canonical SDSS UV/optical template and smaller NIR composites. The sample size enables luminosity-, mass-, and Eddington-ratio-binned stacks that are useful as empirical templates for photo-z, SED modeling, and QSO selection. Continuum slopes, IR/optical ratios, Case-B line ratios, and anti-Baldwin EWs are measured quantities on the delivered stacks rather than model-forced results. The authors openly flag host contamination and recommend high-L sub-composites when host light must be minimized, which strengthens the paper’s utility. If the trends hold under fuller SPHEREx coverage, the work will be a standard reference for AGN continuum and torus studies.

major comments (2)
  1. Sections 4.2 and 5.3 and Figure 6: the luminosity dependence of α_ u,IR and of f_ u(2.4 µm)/f_ u(0.41 µm) is presented as supporting the receding-torus model, yet the same sections attribute the NIR “knee” and part of the slope variation to residual host light and the fact that long-wavelength bins are dominated by lower-z, lower-L objects. The manuscript needs a quantitative bound (e.g., host-subtracted high-L-only stacks, or a simple stellar-template residual estimate) showing that the IR/optical trend survives after host contamination is controlled; without it the physical interpretation remains suggestive rather than demonstrated.
  2. Section 4.3 and Table 3: EWs of Hα, Paβ, and Paα are reported to increase with L_bol (β ≈ 0.12–0.16), contrary to the classical Baldwin effect. Because SPHEREx cannot separate broad and narrow components and Hα is blended with [N II], the claim that the narrow-line Baldwin effect cannot drive the positive trend needs a clearer quantitative argument (e.g., upper limit on the narrow fraction from the SDSS composite or literature). Otherwise the anti-Baldwin result is interesting but not yet secure.
minor comments (5)
  1. Section 4.1: the optical power-law windows (0.15 µm ± 0.01 µm, 0.55 µm ± 0.005 µm) and NIR windows (1.2 and 2.4 µm, width 0.015 µm) should be stated once in a single methods paragraph so that the fits are fully reproducible without hunting through the text.
  2. Figure 2 vs. Vanden Berk et al. (2001): the discrepancy longward of ~0.4 µm is attributed to host contamination and redshift distribution; a short quantitative note (e.g., median f_host or median z in the overlapping bins) would make the comparison more useful.
  3. Table 2 and Appendix tables: only a few rows are shown; ensure the full electronic tables include wavelength, median/mean/geometric-mean fluxes, and N per bin for all subsamples, and that units (arbitrary f_ u) are stated consistently.
  4. Section 5.3: the variability discussion estimates <0.1 mag on ~15-day rest-frame timescales; a brief citation to the SPHEREx sampling cadence or a reference light-curve study would strengthen the claim that variability bias is negligible.
  5. Typographical: “photoionizes” in the Introduction should be “photoionizes” → “photoionizes the dense clouds” is fine but the sentence is slightly incomplete; also “BAL PROB= 0” spacing and “ZWARNING=0” formatting are inconsistent.

Circularity Check

0 steps flagged

No significant circularity: empirical stack of observed SPHEREx fluxes with direct continuum and line measurements.

full rationale

The paper constructs median/mean/geometric-mean composites by iterative normalization and stacking of rest-frame SPHEREx spectrophotometry for ~61 000 SDSS type-1 QSOs (Sections 2–3). Continuum indices α_ν,opt and α_ν,IR are ordinary least-squares power-law fits to fixed, line-free windows; IR/optical flux ratios, Gaussian line fluxes, and EWs are measured after local continuum subtraction (Section 4). Luminosity, mass and Eddington-ratio bins use literature bolometric corrections and virial masses solely for sample division; they do not enter the spectral shape. Comparisons to Case B recombination and the receding-torus model are external literature benchmarks, not inputs that force the measured values. Host-contamination caveats are acknowledged by the authors themselves and do not render any reported quantity tautological. No self-definitional loop, fitted-parameter-as-prediction, load-bearing self-citation uniqueness claim, or renamed known result appears in the derivation chain.

Axiom & Free-Parameter Ledger

4 free parameters · 5 axioms · 0 invented entities

The work is an empirical stacking analysis. It inherits standard cosmological parameters, extinction laws, bolometric corrections and Case-B recombination ratios from the literature; the only free choices that affect the reported numbers are the continuum windows used for power-law fits, the host-fraction cut, and the S/N/channel thresholds that define the sample. No new physical entities are postulated.

free parameters (4)
  • optical continuum windows (0.15 µm ±0.01 µm, 0.55 µm ±0.005 µm) = α_ν,opt = −0.10
    Chosen by hand as relatively line-free regions; the resulting α_ν,opt = −0.10 depends on this choice.
  • NIR continuum windows (1.2 µm and 2.4 µm, width 0.015 µm) = α_ν,IR = −1.46
    Selected to avoid strong emission features and the host-induced knee; yields α_ν,IR = −1.46.
  • host-galaxy fraction cut f_host < 0.1 at 5100 Å = 0.1
    Threshold taken from Wu & Shen (2022) decomposition; directly controls residual stellar contamination that the authors themselves flag as still important in the NIR.
  • median S/N per channel ≥ 2 and ≥ 50 spectral channels = S/N≥2, N_chan≥50
    Quality cuts that reduce the parent sample to ~22 %; affect which luminosity and redshift regimes dominate each rest-frame bin.
axioms (5)
  • domain assumption Standard flat ΛCDM cosmology (H0=70, Ωm=0.3, ΩΛ=0.7) for luminosity distances
    Used only to compute L_bol for binning; stated in the introduction.
  • domain assumption Bolometric corrections of Richards et al. (2006) at 5100, 3000, 1350 Å
    Adopted from Wu & Shen (2022) to define L_bol bins.
  • domain assumption Case B recombination line ratios for typical BLR conditions (Paα/Hα≈0.11, Paβ/Hα≈0.10)
    Used as the benchmark against which measured Paschen-to-Balmer ratios are compared (Section 4.3).
  • domain assumption Galactic extinction law of Gordon et al. (2023) with RV=3.1 and Schlegel/Schlafly maps
    Applied to every spectrum before stacking (Section 2.2).
  • standard math Geometric mean preserves power-law continuum shape
    Standard property of the geometric mean; invoked to justify one of the three stacking estimators.

pith-pipeline@v1.1.0-grok45 · 27000 in / 3402 out tokens · 38503 ms · 2026-07-14T12:56:04.578250+00:00 · methodology

0 comments
read the original abstract

We present a composite spectrum of $\sim 61,000$ type 1 SDSS QSOs (median $z \approx 1.26$), constructed using SPHEREx spectrophotometric data and covering a rest-frame wavelength range of $0.14-4.5~\mu$m. The SPHEREx mission surveys the entire sky in 102 near-infrared spectral channels spanning $0.75-5.0~\mu$m with a spectral resolution of $R \approx 35-130$, providing a unique dataset for building a statistically robust QSO composite. We find that the UV and optical continuum of the resulting composite can be described by a power law, $f_\nu \propto \nu^{\alpha_\nu}$, with a best-fit spectral index of $\alpha_\nu = -0.10$, while the near-infrared continuum is well-fit with a spectral index of $-1.46$. The power-law indices in both the optical and near-infrared regimes strongly depend on properties of QSOs, such that more luminous QSOs tend to exhibit flatter UV/optical and steeper near-infrared continua compared to those of less luminous ones. The IR-to-optical flux ratio decreases with increasing AGN luminosity, consistent with the predictions of the receding torus model. The line ratios of broad emission lines, including H$\alpha$, Pa$\beta$, and Pa$\alpha$, are in good agreement with predictions from Case B recombination, suggesting that internal extinction is almost negligible. The equivalent widths of these emission lines are proportional to AGN luminosity, contrary to the trend expected from the Baldwin effect. Finally, the shape of the composite is sensitive to host-galaxy contamination, which must be considered when utilizing this QSO composite for subsequent scientific applications.

Figures

Figures reproduced from arXiv: 2607.10282 by Andreas L. Faisst, Asantha Cooray, Bomee Lee, Brendan P. Crill, Chi H. Nguyen, Daniel C. Masters, Dohyeong Kim, Howard Hui, Jeong Hwan Lee, Jeonghyun Pyo, Jong-Hak Woo, Kyuseok Oh, Michael Zemcov, Minjin Kim, Olivier Dore, Richard M. Feder, Woong-Seob Jeong, Yi-Kuan Chiang, Yongjung Kim, Yujin Yang, Yun-Ting Cheng, Zhaoyu Huai.

Figure 1
Figure 1. Figure 1: Physical properties of the final QSO sample, used to generate the composite spectra. Redshifts and bolo￾metric luminosities are shown in the top panels, while BH mass and Eddington ratios are displayed in the bottom pan￾els. regimes (W. Ren et al. 2024). We note that while we at￾tempt to minimize host galaxy contamination using the SDSS optical spectra, the host contribution remains sig￾nificant in the SPH… view at source ↗
Figure 2
Figure 2. Figure 2: Composite spectra generated from the SPHEREx dataset. For the top panel, blue, magenta, and red lines denote the composites with median, mean, and geometric mean, respectively. Shaded regions indicate the 16th–84th percentile range of the flux distribution in each spectral bin. Vertical gray lines indicate the rest-frame wavelengths of major emission lines, as labeled at the top of the plot. The green line… view at source ↗
Figure 3
Figure 3. Figure 3: Same as [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Same as [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 6
Figure 6. Figure 6: (a) IR spectral index, (b) flux ratio of IR to optical continuum, and (c) EWs of emission lines in the composites as a function of AGN bolometric luminosity. In the right panel, the dashed lines represent the best-fit linear relations for each emission line. correlation between accretion disk temperature and the accretion rate (e.g., U. Giveon et al. 1999; B. C. Wilhite et al. 2005). While this trend is mo… view at source ↗
Figure 7
Figure 7. Figure 7: Examples of SPHEREx spectra for host-dominated (left) and QSO-dominated (right) objects. The classification is based on the flux ratio f3.7/f1.6, representing the relative contribution of the host galaxy versus the AGN. In the left panel, the spectra are color-coded by their f3.7/f1.6 values (orange, green, magenta, and red, in order of decreasing ratio), highlighting the transition in spectral shape as th… view at source ↗
Figure 8
Figure 8. Figure 8: Examples of spectral line fits for the composite spectrum of the entire sample. The black histogram shows the observed data, while the blue line denotes the best-fit single-Gaussian model. compared to the other works (see also J. Selsing et al. 2016). Furthermore, previous studies generally excluded low￾luminosity AGNs from their analyses. This suggests that their host galaxy contribution is minimal, leadi… view at source ↗
Figure 9
Figure 9. Figure 9: Comparison of NIR QSO composite spectra with previous studies. While the red line denotes the composite spectrum from this study, blue, black, and magenta dashed lines denote the composites from E. Glikman et al. (2006), A. Hern´an-Caballero et al. (2016), and D. Kim et al. (2015), respectively. contamination introduces nonlinear effects along the wavelength axis and systematic biases in the spectral shape… 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

73 extracted references · 11 canonical work pages

  1. [1]

    G., Aguilar, J., Ahlen, S., et al

    Adame, A. G., Aguilar, J., Ahlen, S., et al. 2025, JCAP, 2025, 021, doi: 10.1088/1475-7516/2025/02/021

    doi:10.1088/1475-7516/2025/02/021 2025

  2. [2]

    P., Crill, B

    Akeson, R., Dubois-Felsmann, G. P., Crill, B. P., et al. 2025, arXiv e-prints, arXiv:2511.15823, doi: 10.48550/arXiv.2511.15823 14Kim et al. Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al. 2013, A&A, 558, A33, doi: 10.1051/0004-6361/201322068 Astropy Collaboration, Price-Whelan, A. M., Sip˝ ocz, B. M., et al. 2018, AJ, 156, 123, doi: 10....

    Pith/arXiv arXiv doi:10.48550/arxiv.2511.15823 2025

  3. [3]

    Baldwin, J. A. 1977, ApJ, 214, 679, doi: 10.1086/155294

    doi:10.1086/155294 1977

  4. [4]

    J., Aboobaker, A

    Bock, J. J., Aboobaker, A. M., Adamo, J., et al. 2026, ApJ, 999, 139, doi: 10.3847/1538-4357/ae2be2

    doi:10.3847/1538-4357/ae2be2 2026

  5. [5]

    W., Cheng, L., Shields, G

    Bonning, E. W., Cheng, L., Shields, G. A., Salviander, S., & Gebhardt, K. 2007, ApJ, 659, 211, doi: 10.1086/510712

    doi:10.1086/510712 2007

  6. [6]

    A., & Green, R

    Boroson, T. A., & Green, R. F. 1992, ApJS, 80, 109, doi: 10.1086/191661

    doi:10.1086/191661 1992

  7. [7]

    2023, ApJ, 944, 107, doi: 10.3847/1538-4357/acb3c2

    Chaussidon, E., Y` eche, C., Palanque-Delabrouille, N., et al. 2023, ApJ, 944, 107, doi: 10.3847/1538-4357/acb3c2

    doi:10.3847/1538-4357/acb3c2 2023

  8. [8]

    M., Rhook, K., Corbett, E

    Croom, S. M., Rhook, K., Corbett, E. A., et al. 2002, MNRAS, 337, 275, doi: 10.1046/j.1365-8711.2002.05910.x

    doi:10.1046/j.1365-8711.2002.05910.x 2002

  9. [9]

    B., Hennawi, J

    Davies, F. B., Hennawi, J. F., Ba˜ nados, E., et al. 2018, ApJ, 864, 142, doi: 10.3847/1538-4357/aad6dc

    doi:10.3847/1538-4357/aad6dc 2018

  10. [10]

    E., et al

    Dayal, P., Volonteri, M., Greene, J. E., et al. 2025, A&A, 697, A211, doi: 10.1051/0004-6361/202449331

    doi:10.1051/0004-6361/202449331 2025

  11. [11]

    C., et al

    Dietrich, M., Hamann, F., Shields, J. C., et al. 2002, ApJ, 581, 912, doi: 10.1086/344410

    doi:10.1086/344410 2002

  12. [12]

    Fabian, A. C. 2012, ARA&A, 50, 455, doi: 10.1146/annurev-astro-081811-125521

    doi:10.1146/annurev-astro-081811-125521 2012

  13. [13]

    A., Richards, G

    Fan, X., Strauss, M. A., Richards, G. T., et al. 2006, AJ, 131, 1203, doi: 10.1086/500296

    doi:10.1086/500296 2006

  14. [14]

    J., Hewett, P

    Francis, P. J., Hewett, P. C., Foltz, C. B., et al. 1991, ApJ, 373, 465, doi: 10.1086/170066

    doi:10.1086/170066 1991

  15. [15]

    2024, ApJS, 271, 54, doi: 10.3847/1538-4365/ad2ae6

    Fu, Y., Wu, X.-B., Li, Y., et al. 2024, ApJS, 271, 54, doi: 10.3847/1538-4365/ad2ae6

    doi:10.3847/1538-4365/ad2ae6 2024

  16. [16]

    Giveon, U., Maoz, D., Kaspi, S., Netzer, H., & Smith, P. S. 1999, MNRAS, 306, 637, doi: 10.1046/j.1365-8711.1999.02556.x

    doi:10.1046/j.1365-8711.1999.02556.x 1999

  17. [17]

    J., & White, R

    Glikman, E., Helfand, D. J., & White, R. L. 2006, ApJ, 640, 579, doi: 10.1086/500098

    doi:10.1086/500098 2006

  18. [18]

    D., Clayton, G

    Gordon, K. D., Clayton, G. C., Decleir, M., et al. 2023, ApJ, 950, 86, doi: 10.3847/1538-4357/accb59

    doi:10.3847/1538-4357/accb59 2023

  19. [19]

    J., Forster, K., & Kuraszkiewicz, J

    Green, P. J., Forster, K., & Kuraszkiewicz, J. 2001, ApJ, 556, 727, doi: 10.1086/321600

    doi:10.1086/321600 2001

  20. [20]

    E., & Ho, L

    Greene, J. E., & Ho, L. C. 2005, ApJ, 630, 122, doi: 10.1086/431897 Hern´ an-Caballero, A., Hatziminaoglou, E., Alonso-Herrero, A., & Mateos, S. 2016, MNRAS, 463, 2064, doi: 10.1093/mnras/stw2107

    doi:10.1086/431897 2005

  21. [21]

    Ho, L. C. 2005, ApJ, 629, 680, doi: 10.1086/431643 H¨ onig, S. F., & Beckert, T. 2007, MNRAS, 380, 1172, doi: 10.1111/j.1365-2966.2007.12157.x

    doi:10.1086/431643 2005

  22. [22]

    G., Ross, A

    Hou, J., S´ anchez, A. G., Ross, A. J., et al. 2021, MNRAS, 500, 1201, doi: 10.1093/mnras/staa3234

    doi:10.1093/mnras/staa3234 2021

  23. [23]

    J., Condon, S., et al

    Hui, H., Bock, J. J., Condon, S., et al. 2026, ApJS, 284, 10, doi: 10.3847/1538-4365/ae522c

    doi:10.3847/1538-4365/ae522c 2026

  24. [24]

    S., Netzer, H., et al

    Kaspi, S., Smith, P. S., Netzer, H., et al. 2000, ApJ, 533, 631, doi: 10.1086/308704

    doi:10.1086/308704 2000

  25. [25]

    2010, ApJ, 724, 386, doi: 10.1088/0004-637X/724/1/386

    Kim, D., Im, M., & Kim, M. 2010, ApJ, 724, 386, doi: 10.1088/0004-637X/724/1/386

    work page doi:10.1088/0004-637x/724/1/386 2010

  26. [26]

    H., et al

    Kim, D., Im, M., Kim, J. H., et al. 2015, ApJS, 216, 17, doi: 10.1088/0067-0049/216/1/17

    work page doi:10.1088/0067-0049/216/1/17 2015

  27. [27]

    H., Im, M., Lee, H

    Kim, J. H., Im, M., Lee, H. M., et al. 2012, ApJ, 760, 120, doi: 10.1088/0004-637X/760/2/120

    doi:10.1088/0004-637x/760/2/120 2012

  28. [28]

    H., Im, M., Kim, D., et al

    Kim, J. H., Im, M., Kim, D., et al. 2019, PASJ, 71, 25, doi: 10.1093/pasj/psy144

    work page doi:10.1093/pasj/psy144 2019

  29. [29]

    C., & Im, M

    Kim, M., Ho, L. C., & Im, M. 2006, ApJ, 642, 702, doi: 10.1086/501422

    work page doi:10.1086/501422 2006

  30. [30]

    Kim, M., Son, S., & Ho, L. C. 2024, A&A, 689, A27, doi: 10.1051/0004-6361/202450413

    doi:10.1051/0004-6361/202450413 2024

  31. [31]

    Kim, S., Kim, M., Son, S., & Ho, L. C. 2026, ApJ, 1000, 245, doi: 10.3847/1538-4357/ae4ecd

    work page doi:10.3847/1538-4357/ae4ecd 2026

  32. [32]

    2015, ApJL, 813, L35, doi: 10.1088/2041-8205/813/2/L35

    Kim, Y., Im, M., Jeon, Y., et al. 2015, ApJL, 813, L35, doi: 10.1088/2041-8205/813/2/L35

    work page doi:10.1088/2041-8205/813/2/l35 2015

  33. [33]

    Kormendy, J., & Ho, L. C. 2013, ARA&A, 51, 511, doi: 10.1146/annurev-astro-082708-101811

    doi:10.1146/annurev-astro-082708-101811 2013

  34. [34]

    M., Bock, J

    Korngut, P. M., Bock, J. J., Condon, S., et al. 2026, arXiv e-prints, arXiv:2603.29835, doi: 10.48550/arXiv.2603.29835 Kovaˇ cevi´ c, J., Popovi´ c, L.ˇC., & Dimitrijevi´ c, M. S. 2010, ApJS, 189, 15, doi: 10.1088/0067-0049/189/1/15

    doi:10.48550/arxiv.2603.29835 2026

  35. [35]

    1991, MNRAS, 252, 586, doi: 10.1093/mnras/252.4.586

    Lawrence, A. 1991, MNRAS, 252, 586, doi: 10.1093/mnras/252.4.586

    doi:10.1093/mnras/252.4.586 1991

  36. [36]

    F., et al

    Lusso, E., Worseck, G., Hennawi, J. F., et al. 2015, MNRAS, 449, 4204, doi: 10.1093/mnras/stv516

    doi:10.1093/mnras/stv516 2015

  37. [37]

    W., Higley, A

    Lyke, B. W., Higley, A. N., McLane, J. N., et al. 2020, ApJS, 250, 8, doi: 10.3847/1538-4365/aba623

    doi:10.3847/1538-4365/aba623 2020

  38. [38]

    2013, MNRAS, 430, 3445, doi: 10.1093/mnras/stt143

    Ma, X.-C., & Wang, T.-G. 2013, MNRAS, 430, 3445, doi: 10.1093/mnras/stt143

    doi:10.1093/mnras/stt143 2013

  39. [39]

    2024, ApJ, 971, 75, doi: 10.3847/1538-4357/ad5ce8

    Madau, P., Giallongo, E., Grazian, A., & Haardt, F. 2024, ApJ, 971, 75, doi: 10.3847/1538-4357/ad5ce8

    doi:10.3847/1538-4357/ad5ce8 2024

  40. [40]

    2007, A&A, 468, 979, doi: 10.1051/0004-6361:20077252

    Maiolino, R., Shemmer, O., Imanishi, M., et al. 2007, A&A, 468, 979, doi: 10.1051/0004-6361:20077252

    doi:10.1051/0004-6361:20077252 2007

  41. [41]

    J., & Jarvis, M

    McLure, R. J., & Jarvis, M. J. 2002, MNRAS, 337, 109, doi: 10.1046/j.1365-8711.2002.05871.x

    doi:10.1046/j.1365-8711.2002.05871.x 2002

  42. [42]

    1997, AJ, 114, 228, doi: 10.1086/118467

    Merritt, D. 1997, AJ, 114, 228, doi: 10.1086/118467

    doi:10.1086/118467 1997

  43. [43]

    2020, MNRAS, 499, 210, doi: 10.1093/mnras/staa2780

    Neveux, R., Burtin, E., de Mattia, A., et al. 2020, MNRAS, 499, 210, doi: 10.1093/mnras/staa2780

    doi:10.1093/mnras/staa2780 2020

  44. [44]

    2024, ApJ, 974, 153, doi: 10.3847/1538-4357/ad6e76

    Ren, W., Guo, H., Shen, Y., et al. 2024, ApJ, 974, 153, doi: 10.3847/1538-4357/ad6e76

    doi:10.3847/1538-4357/ad6e76 2024

  45. [45]

    T., Fan, X., Newberg, H

    Richards, G. T., Fan, X., Newberg, H. J., et al. 2002, AJ, 123, 2945, doi: 10.1086/340187 QSO Composite Spectrum from SPHEREx15

    doi:10.1086/340187 2002

  46. [46]

    T., Lacy, M., Storrie-Lombardi, L

    Richards, G. T., Lacy, M., Storrie-Lombardi, L. J., et al. 2006, ApJS, 166, 470, doi: 10.1086/506525

    doi:10.1086/506525 2006

  47. [47]

    E., Ellis, R

    Robertson, B. E., Ellis, R. S., Furlanetto, S. R., & Dunlop, J. S. 2015, ApJL, 802, L19, doi: 10.1088/2041-8205/802/2/L19

    doi:10.1088/2041-8205/802/2/l19 2015

  48. [48]

    P., Myers, A

    Ross, N. P., Myers, A. D., Sheldon, E. S., et al. 2012, ApJS, 199, 3, doi: 10.1088/0067-0049/199/1/3

    doi:10.1088/0067-0049/199/1/3 2012

  49. [49]

    F., & Finkbeiner, D

    Schlafly, E. F., & Finkbeiner, D. P. 2011, ApJ, 737, 103, doi: 10.1088/0004-637X/737/2/103

    doi:10.1088/0004-637x/737/2/103 2011

  50. [50]

    J., Finkbeiner, D

    Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, ApJ, 500, 525, doi: 10.1086/305772

    doi:10.1086/305772 1998

  51. [51]

    Selsing, J., Fynbo, J. P. U., Christensen, L., & Krogager, J.-K. 2016, A&A, 585, A87, doi: 10.1051/0004-6361/201527096

    doi:10.1051/0004-6361/201527096 2016

  52. [52]

    I., & Sunyaev, R

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

    1973

  53. [53]

    Shen, Y., & Ho, L. C. 2014, Nature, 513, 210, doi: 10.1038/nature13712

    doi:10.1038/nature13712 2014

  54. [54]

    T., Strauss, M

    Shen, Y., Richards, G. T., Strauss, M. A., et al. 2011, ApJS, 194, 45, doi: 10.1088/0067-0049/194/2/45

    doi:10.1088/0067-0049/194/2/45 2011

  55. [55]

    B., Horne, K., et al

    Shen, Y., Hall, P. B., Horne, K., et al. 2019, ApJS, 241, 34, doi: 10.3847/1538-4365/ab074f

    doi:10.3847/1538-4365/ab074f 2019

  56. [56]

    2025, ApJ, 985, 107, doi: 10.3847/1538-4357/adc57f

    Shim, H., Baek, J., Kim, D., et al. 2025, ApJ, 985, 107, doi: 10.3847/1538-4357/adc57f

    work page doi:10.3847/1538-4357/adc57f 2025

  57. [57]

    Silk, J., & Rees, M. J. 1998, A&A, 331, L1, doi: 10.48550/arXiv.astro-ph/9801013

    Pith/arXiv arXiv doi:10.48550/arxiv.astro-ph/9801013 1998

  58. [58]

    2005, MNRAS, 360, 565, doi: 10.1111/j.1365-2966.2005.09043.x

    Simpson, C. 2005, MNRAS, 360, 565, doi: 10.1111/j.1365-2966.2005.09043.x

    doi:10.1111/j.1365-2966.2005.09043.x 2005

  59. [59]

    Son, S., Kim, M., & Ho, L. C. 2022, ApJ, 927, 107, doi: 10.3847/1538-4357/ac4dfc

    doi:10.3847/1538-4357/ac4dfc 2022

  60. [60]

    Son, S., Kim, M., & Ho, L. C. 2023, ApJ, 953, 175, doi: 10.3847/1538-4357/ace165

    work page doi:10.3847/1538-4357/ace165 2023

  61. [61]

    C., & Li, R

    Son, S., Kim, M., Ho, L. C., & Li, R. 2025, ApJ, 995, 37, doi: 10.3847/1538-4357/ae1ef1

    work page doi:10.3847/1538-4357/ae1ef1 2025

  62. [62]

    J., Benford, D

    Stern, D., Assef, R. J., Benford, D. J., et al. 2012, ApJ, 753, 30, doi: 10.1088/0004-637X/753/1/30

    doi:10.1088/0004-637x/753/1/30 2012

  63. [63]

    W., Rix, H.-W., et al

    Storey-Fisher, K., Hogg, D. W., Rix, H.-W., et al. 2024, ApJ, 964, 69, doi: 10.3847/1538-4357/ad1328 Vanden Berk, D. E., Richards, G. T., Bauer, A., et al. 2001, AJ, 122, 549, doi: 10.1086/321167

    doi:10.3847/1538-4357/ad1328 2024

  64. [64]

    2002, ApJ, 571, 733, doi: 10.1086/340045

    Vestergaard, M. 2002, ApJ, 571, 733, doi: 10.1086/340045

    doi:10.1086/340045 2002

  65. [65]

    E., et al

    Virtanen, P., Gommers, R., Oliphant, T. E., et al. 2020, Nature Methods, 17, 261, doi: 10.1038/s41592-019-0686-2

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

  66. [66]

    Voit, G. M. 1992, MNRAS, 258, 841, doi: 10.1093/mnras/258.4.841

    doi:10.1093/mnras/258.4.841 1992

  67. [67]

    C., Vanden Berk, D

    Wilhite, B. C., Vanden Berk, D. E., Kron, R. G., et al. 2005, ApJ, 633, 638, doi: 10.1086/430821

    doi:10.1086/430821 2005

  68. [68]

    2022, ApJS, 263, 42, doi: 10.3847/1538-4365/ac9ead

    Wu, Q., & Shen, Y. 2022, ApJS, 263, 42, doi: 10.3847/1538-4365/ac9ead

    doi:10.3847/1538-4365/ac9ead 2022

  69. [69]

    Xie, Y., & Ho, L. C. 2019, ApJ, 884, 136, doi: 10.3847/1538-4357/ab4200

    doi:10.3847/1538-4357/ab4200 2019

  70. [70]

    Xie, Y., & Ho, L. C. 2022, ApJ, 925, 218, doi: 10.3847/1538-4357/ac32e2

    work page doi:10.3847/1538-4357/ac32e2 2022

  71. [71]

    2023, ApJS, 264, 9, doi: 10.3847/1538-4365/ac9ea8

    Yang, Q., & Shen, Y. 2023, ApJS, 264, 9, doi: 10.3847/1538-4365/ac9ea8

    work page doi:10.3847/1538-4365/ac9ea8 2023

  72. [72]

    Zhuang, M.-Y., & Ho, L. C. 2020, ApJ, 896, 108, doi: 10.3847/1538-4357/ab8f2e

    doi:10.3847/1538-4357/ab8f2e 2020

  73. [73]

    C., & Shangguan, J

    Zhuang, M.-Y., Ho, L. C., & Shangguan, J. 2021, ApJ, 906, 38, doi: 10.3847/1538-4357/abc94d 16Kim et al. T able A1.Median SPHEREx QSO Composite Spectra for Subsamples Divided byL bol. bol0 bol1 bol2 bol3 bol4 λrest fν λrest fν λrest fν λrest fν λrest fν 0.2420 0.0926 0.2232 0.2928 0.1967 0.1800 0.1712 0.1981 0.1408 0.3753 0.2475 0.0834 0.2258 0.3503 0.201...

    doi:10.3847/1538-4357/abc94d 2021