Do little red dots really form a distinct class of astronomical objects?

Jean-Baptiste Billand, David Elbaz, Maximilien Franco, Fabrizio Gentile, Emanuele Daddi, Mauro Giavalisco, Dale D. Kocevski, Joseph S. W. Lewis, Benjamin Magnelli, Valentina Sangalli, Maxime Tarrasse

First listed 2026-04-13 | Last updated 2026-04-28

Abstract

JWST observations have identified a class of enigmatic sources known as "Little Red Dots" (LRDs). These have been interpreted as a distinct class of active galactic nuclei (AGN) and host galaxies, potentially involving "quasi-stars" or Black Hole stars (BH*). However, two questions remain: is there a clear discontinuity between LRDs and field galaxies, and do LRDs form a homogeneous population? In this work, we address these issues by introducing a continuous metric to evaluate the "LRDness" of galaxies. We measure their compactness ($δ_{compact}$), the sharpness of the V-shaped spectral energy distribution ($δ_{v-shape}$), and the strength of the broad Balmer line emission. We apply this approach to a sample of ~48,000 galaxies with photometric and ~5,000 with spectroscopic information, selected over ~750 arcmin^2. We find that V-shape prominence correlates strongly with morphology without a clear transition at common LRD selection thresholds: the fraction of compact galaxies rises continuously with V-shape intensity. Similarly, broad H$α$ strength increases with both V-shape sharpness and compactness. The [N II] deficit is not an exclusive feature of LRDs but a global property of compact, metal-poor galaxies. Only the 3% most extreme LRDs present a prominent Balmer break (>3) of potentially non-stellar origin. LRDs and non-LRDs follow similar trends in the evolution of the Balmer decrement with V-shape sharpness, suggesting a shared physical origin, likely dust attenuation. Estimated dust masses (~4-7 x 10^4 M_{sun}) and luminosities are low enough to account for their non-detection by ALMA. We conclude that most LRDs do not represent a separate class of objects, but rather the extreme tail of a continuous distribution of galaxies and broad H$α$ emitters, consistent with a classical broad line region and dust attenuation.

Short digest

This paper asks whether little red dots really form a distinct population or whether they are the extreme end of a broader high-redshift galaxy distribution. The main result is that once color, compactness, and emission-line boosting are examined together, the LRD selection looks more continuous with the parent population than a cleanly separate class. The paper matters because it challenges the assumption that LRDs must correspond to a unique physical channel.

Key figures to inspect

• Figure 2 is the must-see figure: it places the sources in the color plane and shows how dust, stellar evolution, and the published LRD cuts relate to one another, which is central to the paper's claim that LRDs may not form a distinct class.
• Figure 4 is the population-comparison figure: it shows where the known LRD sample sits relative to the broader galaxy population in compactness-color space.
• Figure 1 is a useful bias check: it quantifies the H-alpha boosting effect in PRISM colors and shows that the strongest contamination is limited to a minority of the brighter galaxies.
• Figure 3 is the survey-level diagnostic, showing how extreme V-shaped SEDs and compactness are distributed across the fields used in the analysis.