Local Analogs of Little Red Dots: Clues to the X-ray Weakness

Abstract

A major question about JWST's little red dots (LRDs) at $z\sim4-8$ is the origin of their apparent X-ray weakness. Even deep Chandra stacks yield only tentative detections or broad-band limits and cannot distinguish among super-Eddington accretion, intrinsic X-ray weakness, and Compton-thick (CT) obscuration. Local analogs offer a powerful complementary route. At $z\sim0.3$, similarly luminous sources require $\sim10^{3-4}$ less exposure for comparable signal-to-noise, suffer less source confusion, and directly probe the diagnostic rest-frame soft X-rays. Here we present a pilot Einstein Probe/FXT study of six "local little red dots" (LLRDs), V-shaped broad-line green pea galaxies selected as low-redshift counterparts of LRDs. These sources show X-ray diversity: "X-normal" LLRDs resemble type 1 AGNs in X-ray luminosity and optical variability, whereas the "X-weak" stack lies $\sim0.5-0.7$ dex below standard unobscured-quasar scaling relations while remaining compatible with the representative LRD stack and 95% of published individual limits. The X-weak LLRDs also lack detected optical variability and have redder optical continua than X-normal LLRDs, making their V-shaped SEDs closer to those of high-redshift LRDs. The detection of the X-weak stack at $\sim3.3σ$ with the soft-X-ray-sensitive EP-FXT provides rest-frame $\sim2-3$ keV constraints difficult to reconcile with a typical unobscured super-Eddington model or a local-AGN-like CT geometry examined here, although modified variants remain possible. The simplest alternative is a mildly X-ray-weak corona ($Δα_{\mathrm{ox}}\sim-0.1$ to $-0.2$) with Compton-thin absorption ($N_{\mathrm{H}}\sim10^{22-23}$ cm$^{-2}$), consistent with the independent Balmer decrement/break diagnostics. Together with growing samples of local LRD analogs, our results offer a practical path for testing the physical nature of LRD X-ray weakness.