Double Dots: Compact Pairs Mark Little Red Dots and High-Redshift Broad-line AGNs

A. J. Barger, L. L. Cowie

First listed 2026-05-28 | Last updated 2026-05-27

Abstract

We utilize JWST imaging of the massive lensing cluster field A2744 to find close pairs of compact sources with separations less than 0.25". A large fraction of these "Double Dots" correspond to Little Red Dots (LRDs) or high-redshift broad-line active galactic nuclei (BLAGNs). Our analysis of 31 identified pairs reveals a median separation of 0.15". Statistical comparison against a uniform background shows that these are mostly physical pairs. We find that at least 16 of the 24 previously published LRDs in this field (~67%) are such pairs, as are both of the high-redshift BLAGNs. We demonstrate that the presence of a companion can significantly contaminate the measured spectral energy distribution, potentially masking the characteristic ``v-shape" used for LRD classification. Furthermore, our 2D spectroscopic analysis of several pairs reveals that BLAGN activity is not confined to the redder member of the pair but can originate in either one. Since most LRDs contain broad emission lines, our findings suggest that close pairs are extremely effective markers of galaxies with broad lines at high redshift. We speculate on possible mechanisms, concluding that we are likely seeing merger-driven accretion.

Short digest

Using JWST imaging of the lensed A2744 field, this paper defines a sample of 31 sub-0.25 arcsecond compact pairs, or “Double Dots,” and shows their separations are strongly skewed to small values, consistent with most being physical systems rather than chance alignments. The main empirical result is that close pairs are extremely common among known broad-line high-redshift sources in the field: at least 16 of 24 published little red dots and both previously identified high-redshift BLAGNs are double systems, with a median observed separation of 0.15 arcseconds. The authors then show that unresolved companions can materially reshape the observed SED, including washing out the canonical LRD v-shape, and that 2D spectroscopy can place the broad-line emission on either the redder or bluer member of the pair. That makes compact pairing itself a powerful marker of broad-line activity at high redshift and points to merger-driven accretion as a plausible underlying trigger.

Key figures to inspect

• Figure 2. Use this as the core selection-validation figure. It shows that the pair-separation distribution rises steeply toward small offsets and is inconsistent with a uniform background, which is the paper’s key evidence that the Double Dots are mostly physical pairs rather than projected neighbors.
• Figure 5. This is the most efficient visual summary of what the authors are actually finding in the imaging: multiple literature LRDs in A2744 resolved into a red compact source plus a nearby blue companion within the search radius. It makes the paper’s claim concrete and immediately shows why treating these systems as single objects can bias classification.
• Figure 8. Recommend this for the SED-contamination result. The decomposed photometry of DD30 and DD31 clearly illustrates how one component can dominate the rest-frame UV while the other dominates the rest-frame optical, demonstrating why blended catalog fluxes can hide or distort the v-shaped SED used in LRD selection.
• Figure 9. This is the sharpest figure for the spectroscopic reassignment argument. The 2D spectra show that the broad-line emission in the DD58 and DD59 pair is associated with the bluer companion rather than the red LRD candidate, directly supporting the paper’s conclusion that BLAGN activity is not confined to the redder member.
• Figure 12. Use this as the later-stage synthesis figure because it connects the pair population to a physical interpretation. By showing that most systems with broad lines or v-shapes lie among the more strongly bound pairs, it is the clearest captioned diagnostic linking compact pairing to the environments most relevant for AGN triggering and the merger-driven accretion picture.