This paper asks whether tiny, very dense galaxies that form early in the Universe can produce the strange signals seen in “Little Red Dots” (LRDs) without needing an active galactic nucleus (AGN). LRDs are compact sources seen by JWST that show a blue ultraviolet (UV) light, a red optical light, and broad hydrogen emission lines called Balmer lines. The authors use high-resolution cosmological simulations and radiative transfer to make mock observations and compare them directly to LRD measurements.

The team used the FIRE-2 galaxy-formation simulations run with the GIZMO code. These simulations include gas cooling, star formation, and stellar feedback. They then applied three-dimensional dust radiative transfer and produced synthetic emission-line data cubes to predict both continuum and line signatures. In the simulated progenitors of present-day group-sized halos, they found short-lived phases at redshift z ≈ 4–8 (lasting about 150–400 million years) when strong gas inflows build very compact stellar cores. These cores have stellar masses around 10^8.5–10^10.5 solar masses, effective radii below ~300 parsecs, UV magnitudes −23 to −20, and very high circular velocities above 500 km s^−1.

Several observed LRD features are reproduced by these compact stellar systems. The simulations give strong Balmer breaks (a drop in continuum flux across certain optical wavelengths that often signals a population of stars), blue UV slopes (βUV ≈ −1.25), dust masses consistent with some limits, non-detections in ALMA submillimeter data, and Balmer-line widths up to about 1500 km s^−1 produced by galaxy-scale motions. In other words, a very dense stellar core can explain the UV light, a substantial Balmer break, and intermediate broadening of Balmer lines in some LRDs.

There are clear limits to the stellar-only explanation. The simulated galaxies do not reproduce the red rest-optical continuum seen in many LRDs, nor the most extreme Balmer breaks (greater than about 2.5) or the very broad lines above ~2000 km s^−1. They also do not show the spectroscopic pattern expected from a broad-line AGN (broad Balmer lines together with narrow forbidden lines). The simulations further show that these compact conditions can feed gas efficiently to central black holes, so a hybrid picture is plausible: compact stars could set the UV continuum and intermediate line widths while a central AGN might add the red optical light and the most extreme line signatures.