Between Degeneracy and Evolution: UV-to-optical Insights into the BH$^*$ Model in Little Red Dots

Rosa M. Mérida, Marcin Sawicki, Chris J. Willott, Gaia Gaspar, Kartheik G. Iyer

First listed 2026-06-11 | Last updated 2026-06-10

Abstract

Little Red Dots (LRDs) are a heterogeneous class of objects, with several proposed scenarios for their physical nature and evolution. While these theories have been tested on individual LRDs using limited spectral features, a systematic Bayesian analysis of the LRD population incorporating the different models across a broad wavelength range is still lacking. In this study, we conduct a consistent ultraviolet (UV)-to-optical continuum fitting analysis of 66 LRDs at 2<z<6 using JWST/NIRSpec PRISM spectroscopy. Employing a modified version of Bagpipes--including blackbody (BB) emission affected by Balmer absorption, stellar and nebular emission attenuated by dust, and an active galactic nucleus (AGN) component--we assess the performance of the black hole star (BH*) model in describing the LRD population. We adopt broad priors and therefore do not impose any specific physical scenario. Our results show that only ~6% of LRDs with statistically robust solutions (52 objects in total) are best-fit by a BH* in the optical and a host galaxy in the UV. ~8% of LRDs show BB-dominated optical continua but lack a stellar component or exhibit AGN UV leakage. Most LRDs are dominated by stellar and/or AGN emission in the optical, with minor BB contribution. When we adopt a prior that disfavors a strong AGN continuum to enforce BH*-like solutions, the percentage of BH$^*$ systems increases to ~40%, highlighting the strong degeneracy between a BH* solution and alternative scenarios. Even when BH*-like solutions are enforced, many LRDs still require a stellar-dominated optical continuum. This may reveal limitations of the BH* model or point to an evolutionary sequence in which the BB contribution decreases as the host grows, leading to lower BB temperatures and higher stellar masses at lower z. In this scenario, more pronounced ''V'' shapes would correspond to later stages in LRD evolution.

Short digest

This paper fits the UV-to-optical continua of 66 JWST/NIRSpec PRISM little red dots at 2<z<6 with a modified Bagpipes model that allows blackbody, stellar plus nebular, dust, and AGN components to compete instead of hard-wiring a BH* interpretation. ([arxiv.org](https://arxiv.org/abs/2606.12355)) Only about 6% of the 52 objects with statistically robust solutions end up in the clean BH* configuration of a BB-dominated optical continuum plus a host-galaxy UV component, while another ~8% have BB-dominated optical light but either no stellar UV component or signs of AGN UV leakage; most LRDs instead require stellar and or AGN emission in the optical. ([arxiv.org](https://arxiv.org/abs/2606.12355)) When the priors are changed to suppress a strong AGN continuum, the BH*-like fraction jumps to about 40%, so the central result is that BH* solutions are strongly degenerate with more conventional mixed stellar and AGN scenarios rather than uniquely preferred across the population. ([arxiv.org](https://arxiv.org/abs/2606.12355)) The authors therefore leave the sample pointing either to shortcomings in the current BH* picture or to an evolutionary sequence in which the BB component weakens and cools as the host grows, with sharper V-shaped continua marking later stages. ([arxiv.org](https://arxiv.org/abs/2606.12355))

Key figures to inspect

• Figure 1. Use this as the setup figure if it shows the LRD selection definition in rest-frame UV and optical slope space and the comparison to earlier spectroscopic samples. It is the clearest place to establish what objects enter the 66-source parent analysis and why the paper treats them as V-shaped continuum-selected LRDs rather than a generic PRISM sample.
• Figure 5. This is a strong candidate for the main evidence figure if it presents the demographic split of continuum solutions under the broad-prior fit. The key lesson to pull out is that the clean BH* class is a small minority, while stellar-dominated and or AGN-contaminated optical continua account for most of the population.
• Figure 10. Recommend this if it is the figure that isolates the statistically robust subsample or visualizes where BB-dominated but non-BH* solutions sit. It matters because the headline percentages are defined only after restricting to robust fits, and this is where AGN UV leakage or the absence of a needed stellar UV component becomes a concrete diagnostic rather than an abstract caveat.
• Figure 12. This is likely one of the most important later-paper figures if it compares the fiducial broad-prior results to the BH*-enforced prior experiment. It should be used to show that the inferred BH* fraction can be driven from about 6% to about 40% by prior choice alone, which is the paper's clearest demonstration of model degeneracy.
• Figure 14. Prioritize this as a synthesis figure if it links BB temperature, host stellar mass, redshift, or V-shape strength in the enforced or combined analysis. This is the figure most likely to carry the paper's evolutionary reading, namely that lower-redshift systems host cooler and weaker BB components in more massive galaxies, with more pronounced V-shapes tracing later stages.