2605.04685v1
Compact, AGN-hosting Dwarf Galaxies with "Little Red Dots"-like SEDs in the Local Universe
Digest
Bao et al. compile a large local comparison set of 1,204 AGN-hosting dwarf galaxies, matched in broad luminosity regime to JWST little red dots, and classify them with K-means using rest-UV/optical SED shape plus galaxy size. The four resulting groups trace a sequence in UV-optical slope, metallicity, star formation, and dust emission, with roughly half of the local dwarf AGN sample showing the same kind of "V-shaped" SEDs and relatively compact morphologies that define LRDs. But the central result is negative in the most interesting way: the local V-shaped, compact systems are still systematically larger and show different ionization states than z>5 LRDs, implying they are more evolved systems rather than straightforward nearby counterparts. That matters because it argues against a simple local-analog picture and points instead to different black-hole/galaxy growth pathways across cosmic time.
Key figures to inspect
- Figure 2 is the key comparison plot: inspect where the local ADGs fall in UV and optical continuum-slope space relative to the LRD density contours and the Kocevski et al. V-shaped selection window to see how much overlap is real versus superficial.
- Figure 3a should show the paper's main phenomenology most clearly: compare the stacked SEDs of the four K-means groups to the LRD template to identify which groups truly reproduce the V-shaped UV-to-optical behavior and how the UV slope appears to drive the sequence.
- Figure 3b is where the local-analog claim is tested morphologically: use the effective-radius distributions to see whether the V-shaped groups are genuinely compact and how their sizes still remain larger than the high-z LRD expectation.
- Figure 4 helps connect the SED classes to dustier mid-IR behavior: inspect whether the groups with redder optical continua also move to distinct WISE color space, clarifying how AGN-heated dust emission differs across the sequence.
- Figure 1 is worth checking for sample context before over-interpreting the analogy: the redshift distributions by selection channel show how heterogeneous the local ADG compilation is, including the systematically higher-redshift DESI broad-line subset.