The Large-scale Environments of Low-luminosity AGNs at 3.9< z<6 and Implications for Their Host Dark Matter Halos from a Complete NIRCam Grism Redshift Survey Xiaojing Lin1,2aa, Xiaohui Fan2aa, Fengwu Sun3aa, Junyu Zhang2aa, Eiichi Egami2aa, Jakob M. Helton2aa, Feige Wang4aa, Haowen Zhang2aa, Andrew J. Bunker5aa, Zheng Cai1aa, Zhiyuan Ji2aa, Xiangyu Jin2aa, Roberto Maiolino6,7,8aa, Maria Anne Pudoka2aa, Pierluigi Rinaldi2aa, Brant Robertson9aa, Sandro Tacchella6,10aa, Wei Leong Tee2aa, Yang Sun2aa, Christopher N. A. Willmer2aa, Chris Willott11aa, and Yongda Zhu2aa 1 Department of Astronomy, Tsinghua University, Beijing 100084, People’s Republic of China; xiaojinglin.astro@gmail.com 2 Steward Observatory, University of Arizona, 933 N Cherry Avenue, Tucson, AZ 85721, USA 3 Center for Astrophysics | Harvard & Smithsonian, 60 Garden Street, Cambridge, MA 02138, USA 4 Department of Astronomy, University of Michigan, 1085 South University Avenue, Ann Arbor, MI 48109, USA 5 Department of Physics, University of Oxford, Denys Wilkinson Building, Keble Road, Oxford OX13RH, UK 6 Kavli Institute for Cosmology, University of Cambridge, Madingley Road, Cambridge, CB3 0HA, UK 7 Cavendish Laboratory—Astrophysics Group, University of Cambridge, 19 JJ Thomson Avenue, Cambridge, CB3 0HE, UK 8 Department of Physics and Astronomy, University College London, Gower Street, London WC1E 6BT, UK 9 Department of Astronomy and Astrophysics, University of California, Santa Cruz, 1156 High Street, Santa Cruz, CA 95064, USA 10 Cavendish Laboratory, University of Cambridge, 19 JJ Thomson Avenue, Cambridge, CB3 0HE, UK 11 NRC Herzberg, 5071 West Saanich Road, Victoria, BC V9E 2E7, Canada Received 2025 May 5; revised 2025 November 2; accepted 2025 November 10; published 2026 January 14 Abstract We study the large-scale environments and clustering properties of 28 low-luminosity active galactic nuclei (AGNs) at z = 3.9–6 in the GOODS-N field. Our sample, identified from the JWST NIRCam Imaging and WFSS data in Complete NIRCam Grism Redshift Survey and First Reionization Epoch Spectroscopically Complete Observations surveys with either broad Hα emission lines or V-shaped continua, are compared to 782 Hα emitters (HAEs) selected from the same data. These AGNs are located in diverse large-scale environments and do not preferentially reside in denser environments compared to HAEs. Their overdensity field, δ, averaged over (15 h−1 cMpc)3, ranges from −0.56 to 10.56, and shows no clear correlation with broad-line luminosity, black hole (BH) masses, or the AGN fraction. It suggests that >10 cMpc structures do not significantly influence BH growth. We measure the two-point cross-correlation function of AGNs with HAEs, finding a comparable amplitude to that of the HAE autocorrelation. This indicates similar bias parameters and host dark matter halo masses for AGNs and HAEs. The correlation length of field AGNs is 4.26 h−1 cMpc and 7.66 h−1 cMpc at 3.9 < z < 5 and 5 < z < 6, respectively. We infer a median host dark matter halo mass of M Mlog 11.0 11.2h( )/ and host stellar masses of M Mlog 8.4 8.6( ) –/ by comparing with the UNIVERSEMACHINE simulation. Our clustering analysis suggests that low-luminosity AGNs at high redshift reside in normal star-forming galaxies with overmassive BHs. They represent an intrinsically distinct population from luminous quasars and could be a common phase in galaxy evolution. Unified Astronomy Thesaurus concepts: Supermassive black holes (1663); Clustering (1908); Galaxy dark matter halos (1880); Active galactic nuclei (16); Low-luminosity active galactic nuclei (2033); James Webb Space Telescope (2291) 1. Introduction Over the past few decades, the discovery of UV-luminous quasars at z ≳ 6 has reshaped and challenged our under- standing of supermassive black holes (SMBHs) in the early Universe (e.g., E. Bañados et al. 2018; J. Yang et al. 2020; F. Wang et al. 2021; X. Fan et al. 2023). These z > 6 quasars, hosting SMBHs exceeding 109M⊙ (e.g., J. Yang et al. 2021; E. P. Farina et al. 2022), have motivated extensive work on models of BH growth modes and seeding mechanisms (e.g., K. Inayoshi et al. 2022; V. Cammelli et al. 2024; J. Regan & M. Volonteri 2024). Multiple approaches have been proposed to constrain quasar lifetimes and duty cycles, including studies of quasars’ proximity zones (A.-C. Eilers et al. 2017; F. B. Davies et al. 2020; A.-C. Eilers et al. 2020), damping wing features (F. B. Davies et al. 2019; J. F. Hennawi et al. 2024), and clustering properties (A.-C. Eilers et al. 2024; E. Pizzati et al. 2024). These independent approaches suggest quasar lifetimes of ∼106 yr, with low duty cycles (≪1). These findings imply a very short timescale for quasar accretion and extremely rapid BH growth with low radiative efficiency (≲0.1%; F. B. Davies et al. 2019; A.-C. Eilers et al. 2024). Alternatively, the bulk of SMBH growth history might have been enshrouded by dust and thus missed by current surveys (A. Comastri et al. 2015; Y. Ni et al. 2020; E. Lambrides et al. 2024a). The demographics of a broader BH population is the key to understanding the early BH assembly history. The unprecedented near-infrared capabilities of the James Webb Space Telescope (JWST; J. P. Gardner et al. 2023) have brought new insights. Since its launch, JWST has revealed new populations of low-luminosity active galactic nuclei (AGNs) at z > 4, many of which are nicknamed as “little red dots” due to their compact morphology and unique spectral energy distributions (SEDs) in the near-IR wavelengths (e.g., The Astrophysical Journal, 997:61 (15pp), 2026 January 20 https://doi.org/10.3847/1538-4357/ae1eef © 2026. The Author(s). Published by the American Astronomical Society. aaaaaaa Original content from this work may be used under the terms of the Creative Commons Attribution 4.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. 1 https://orcid.org/0000-0001-6052-4234 https://orcid.org/0000-0003-3310-0131 https://orcid.org/0000-0002-4622-6617 https://orcid.org/0000-0002-1574-2045 https://orcid.org/0000-0003-1344-9475 https://orcid.org/0000-0003-4337-6211 https://orcid.org/0000-0002-7633-431X https://orcid.org/0000-0002-4321-3538 https://orcid.org/0000-0002-8651-9879 https://orcid.org/0000-0001-8467-6478 https://orcid.org/0000-0001-7673-2257 https://orcid.org/0000-0002-5768-738X https://orcid.org/0000-0002-4985-3819 https://orcid.org/0000-0003-4924-5941 https://orcid.org/0000-0002-5104-8245 https://orcid.org/0000-0002-4271-0364 https://orcid.org/0000-0002-8224-4505 https://orcid.org/0000-0003-0747-1780 https://orcid.org/0000-0001-6561-9443 https://orcid.org/0000-0001-9262-9997 https://orcid.org/0000-0002-4201-7367 https://orcid.org/0000-0003-3307-7525 mailto:xiaojinglin.astro@gmail.com http://astrothesaurus.org/uat/1663 http://astrothesaurus.org/uat/1908 http://astrothesaurus.org/uat/1880 http://astrothesaurus.org/uat/1880 http://astrothesaurus.org/uat/16 http://astrothesaurus.org/uat/2033 http://astrothesaurus.org/uat/2291 http://astrothesaurus.org/uat/2291 https://doi.org/10.3847/1538-4357/ae1eef https://crossmark.crossref.org/dialog/?doi=10.3847/1538-4357/ae1eef&domain=pdf&date_stamp=2026-01-14 https://creativecommons.org/licenses/by/4.0/ H. B. Akins et al. 2024; J. E. Greene et al. 2024; V. Kokorev et al. 2024; X. Lin et al. 2024; J. Matthee et al. 2024; P. G. Pérez-González et al. 2024; P. Rinaldi et al. 2024). They often exhibit broad Balmer emission lines with FWHM exceeding 1000 km s−1. Assuming these broad emission lines originate from the AGN broad-line region, they suggest SMBHs with masses ranging from 106M⊙ to ≲109M⊙ (e.g., J. E. Greene et al. 2024; X. Lin et al. 2024). While JWST has revealed these previously unseen objects, their nature and connection to UV-luminous quasars remain elusive. For instance, there is an ongoing debate regarding the origin of the so-called “V-shaped” continua observed in many of these objects, which exhibit a reddened rest-frame optical continuum and a bluer UV continuum slope (e.g., Z. Li et al. 2024; Y. Ma et al. 2024). It is also difficult to explain the absence of a hot dust torus (C. M. Casey et al. 2024; P. G. Pérez-González et al. 2024; C. C. Williams et al. 2024), the weakness of X-rays (R. Maiolino et al. 2024; M. Yue et al. 2024a), and the lack of variability in both UV and optical bands (M. Kokubo & Y. Harikane 2025; W. L. Tee et al. 2024; Z. Zhang et al. 2024), which are typical features of UV-luminous quasars. The high spatial resolution of NIRCam short-wavelength imaging reveals that ∼30% of LRDs exhibit signs of interaction, suggesting a potential link to merger-driven activities (P. Rinaldi et al. 2024). Some theoretical hypotheses propose that the broad emission lines may not originate from SMBHs, but rather from unusual galaxy kinematics or outflows (J. F. W. Baggen et al. 2024; M. Kokubo & Y. Harikane 2025). In this paper, we refer to these objects (broad Balmer line emitters or “little red dots”) as low-luminosity AGNs, as they literally exhibit the defining characteristics of AGNs— active and compact. While we assume the broad lines originate from BHs, as is commonly done in other works, we note the possibility of their non-BH origins. The measurements presented in this paper, and comparisons with those of their coeval galaxy populations, are independent of the assumptions of the origins of their energy source in most cases. To understand the nature of low-luminosity AGNs, it is crucial to constrain their large-scale environments and host dark matter halos within the context of cosmological structure formation models. Currently, our knowledge about the host galaxies of low-luminosity AGNs largely comes from SED modeling. However, this method is limited by assumptions about the intrinsic characteristics, leading to significant degeneracy (H. B. Akins et al. 2024; G. C. K. Leung et al. 2024; Y. Ma et al. 2024). It can result in a wide range of stellar masses, from extremely massive hosts (∼1011M⊙) to typical star-forming galaxies (∼108M⊙), depending on how AGN and galaxy light contributions are tuned (e.g., B. Wang et al. 2024). The clustering of low-luminosity AGNs can break this degeneracy by directly comparing their large-scale environ- ments and dark matter halo masses with those of galaxies of similar stellar masses. Clustering analysis also provides insights to bridge the gap between low-luminosity AGNs and UV-bright quasars. UV-luminous quasars are character- ized by strong large-scale clustering over a wide redshift range, consistent with host halo masses of ∼1012.5M⊙ at z > 6 (e.g., H. Chen et al. 2022; J. Arita et al. 2023; A.-C. Eilers et al. 2024; M. Pudoka et al. 2024). If low-luminosity AGNs represent the dust-obscured phase of UV-luminous quasars (e.g., J. Lyu et al. 2024), they should inhabit similar massive dark matter halos and share comparable clustering properties. Observational studies of AGN environments and clustering properties can also inform theoretical models. For example, super-Eddington accretion has been proposed to explain some characteristics of low-luminosity AGNs, such as the V-shaped SEDs, X-ray weaknesses, and lack of variability (K. Inayoshi et al. 2024; M. Kokubo & Y. Harikane 2025; E. Lambrides et al. 2024b; A. Trinca et al. 2024). Super-Eddington accretion is predicted to occur only in low-mass hosts (∼1010M⊙) where the AGN duty cycle is low (<1%; E. Pizzati et al. 2024a). Recently, J.-T. Schindler et al. (2024) reported a low- luminosity AGN in an overdensity, with a minimum host halo mass of 1012.3M⊙. However, systematic studies involving a larger sample of AGNs are still required. J. Matthee et al. (2024b) examined the environments within 1 cMpc of six AGNs in the A2744 lensing cluster field. Nonetheless, nonlinear physical processes may dominate on such small scales (Y. Harikane et al. 2016; T. Herard-Demanche et al. 2023). Their analysis does not extend to larger scales due to survey volume limitations. A clustering analysis on scales ≳10 comoving Mpc (cMpc) is crucial to probe the linear two- halo term and achieve more accurate constraints on halo masses (e.g., J. Arita et al. 2024). In this context, the JWST/NIRCam Wide Field Slitless Spectroscopy (WFSS; T. P. Greene et al. 2016; M. J. Rieke et al. 2023) provides a unique and powerful tool for clustering analyses of low-luminosity AGNs. The JWST/NIRCam WFSS has demonstrated great efficiency in studying high- redshift emission-line galaxies. It is highly effective in identifying broad-line AGNs because of its relatively high spectral resolution, and in probing large-scale structures because of its large field of view (FOV; e.g., F. Sun et al. 2022; J. M. Helton et al. 2024; X. Lin et al. 2024). The selection function is simpler to model compared to the spectroscopic observations with pre-selection (e.g., through JWST NIRSpec). In this work, we compile AGNs and galaxies from a Complete NIRCam Grism Redshift Survey (CON- GRESS) in the GOODS-N field. The dataset includes grism observations from the Cycle-1 program “First Reionization Epoch Spectroscopically Complete Observations” (FRESCO, GO-1895, PI: Oesch; P. A. Oesch et al. 2023) and the Cycle-2 program CONGRESS (GO-3577, PI Eiichi & Sun; F. Sun et al. 2025, in preparation). The two programs target the same area using the F444W and F356W grisms, respectively, together covering a total wavelength range of 3.1–5.0 μm. We collect a large number of low-luminosity AGNs at 4 ≲ z < 6 through their broad Hα emission lines and broadband photometric properties, and simultaneously map their surrounding environments with Hα emitters (HAEs) at the same redshift. The paper is organized as follows. In Section 2, we introduce the datasets and methods to select galaxies and AGNs. In Section 3, we explore the large-scale environments of low- luminosity AGNs. A clustering analysis based on two-point correlation functions is presented in Section 4. Finally, in Section 5, we discuss the potential implications of our measurements for AGN evolution. Throughout this work, a flat ΛCDM cosmology is assumed, with H0 = 70 km s−1 Mpc−1, ΩΛ,0 = 0.7, and Ωm,0 = 0.3. We define h = H0/100 = 0.7. 2 The Astrophysical Journal, 997:61 (15pp), 2026 January 20 Lin et al. 2. Data and Sample 2.1. Imaging and Photometric Catalog We use the images and photometric catalog in the GOODS- N field from the JADES Data Release 312 (F. D’Eugenio et al. 2024). The JADES JWST/NIRCam images in the GOODS-N field include observations from GTO programs 1181 (PI Eisenstein) and GO programs 1895 (FRESCO, PI Oesch), spanning the F090W, F115W, F150W, F182M, F200W, F210M, F277W, F335M, F356W, F410M, and F444W filters. The final mosaics cover from 56 arcmin2 in F090W to 83 arcmin2 in F444W (F. D’Eugenio et al. 2024). We refer to D. J. Eisenstein et al. (2023) for a detailed description of the JADES survey design, and M. J. Rieke et al. (2023) and B. Robertson et al. (2025, in preparation) for the imaging data reduction. For the direct images used for the grism spectra, the JADES images achieve a 5σ point-source depth of 29.38 mag in F444W and 29.97 mag in F356W, with aperture-corrected photometry using an r = 0.15 circular aperture. The JADES GOODS-N photometric catalog includes multiband photometry in the 11 JWST/NIRCam filters as mentioned above, and five HST/ACS filters (F435W, F606W, F775W, F814W, and F850LP). The HST/ACS photometry is based on images from the Hubble Legacy Fields project (G. Illingworth 2017). We refer to B. Robertson et al. (2024) for detailed source detection and photometry measurement meth- ods. The photometric redshifts are estimated utilizing the r = 0.1 circular aperture photometry, using EAZY (G. B. Bram- mer et al. 2008) with galaxy SED templates optimized for high- redshift sources (K. N. Hainline et al. 2024). 2.2. JWST/NIRCam Grism Spectroscopy JWST/NIRCam WFSS observations of the GOODS-N field were obtained in both F356W and F444W filters. The Cycle-1 program, FRESCO, covers 62 arcmin2 of the GOODS-N field through the F444W filter and row-direction grism (Grism R). The FRESCO observations in GOODS-N were split into eight pointings, and the exposure time is 8 × 880 s per pointing. The Cycle-2 program, CONGRESS, targets the same areas in the GOODS-N field observed by FRESCO. CONGRESS adopts the F356W filter and Grism R. CONGRESS includes 12 pointings, and the exposure time is 8× 472 s per pointing. The combination of F356W and F444W grism observations results in a total wavelength coverage at 3.1–5.0 μm. The overlap area between JADES, FRESCO, and CONGRESS is about 62 arcmin2, as illustrated in the bottom panel of Figure 1. The grism data reduction and spectral extraction are detailed in F. Sun et al. (2025, in preparation). Here we briefly summarize the procedures. For individual exposures of grism data and their corresponding short-wavelength (SW) direct images, we performed flat-fielding, subtracted a sigma-clipped median sky background, and aligned the World Coordinate System frames. We then measured the astrometric offsets between the direct images and the JADES DR3 GOODS-N catalog. The offsets were added to the spectral tracing model for accurate wavelength calibration. The spectral tracing and wavelength calibration are based on the Commissioning, Cycle-1, and Cycle-2 calibration data taken in the SMP- LMC-58 field up to June 2024 (PID 1076, 1479, 1480, and 4449; see F. Sun et al. 2023).13 The flux calibration is based on Cycle-1 calibration data (PID: 1076, 1536, 1537, 1538). We optimally extracted the 1D spectra based on the source morphology (K. Horne 1986). 2.3. Hα Emitters at 3.9 < z < 6 We refer readers to X. Lin et al. (2025) for detailed descriptions of the selection procedure for line emitters. In brief, we use 51 pixel median filtering to remove the continua, extract line-only spectra, and detect emission lines in both 1D and 2D spectra. We have developed a semiautomated algorithm to identify emission-line galaxies across the redshift range 0 < z < 9 using the photometric redshifts in the JADES catalog as priors. As a result, we selected 936 HAEs across 3.75 < z < 6.6 with F356W and F444W Grism R in the overlapping regions of the CONGRESS and JADES imaging footprint. This work focuses on a subsample of these HAEs within the redshift range of 3.9 < z < 6. We measured the line flux by fitting the emission line with Gaussian models convolved with the NIRCam grism line spread functions as calibrated by F. Sun et al. (2025, in preparation). The emitter 12h37m30s 00s 36m30s 00s 62°18' 15' 12' 09' RA DE C WFSS-selected AGNs NIRSpec-selected AGNs V-shape-selected AGNs 4.0 4.5 5.0 5.5 6.0 Redshift 0 2 4 N pe r b in WFSS-selected NIRSpec-selected V-shape-selected Figure 1. Top panel: the redshift distributions of AGNs in the GOODS-N field. The red, orange, and yellow histograms represent WFSS-selected, NIRSpec-selected, and V-shaped-selected samples, respectively. Bottom panel: the spatial distribution of AGNs. The blue shaded regions are the F356W WFSS (GO-3577, CONGRESS) footprint, and the black lines enclose the F444W image footprint of JADES DR3 (including FRESCO GO-1895). AGNs at 3.9 < z < 6 are marked as red dots, orange squares, and yellow hexagons for WFSS-selected, NIRSpec-selected, and V-shaped selected sources, respectively. 12 https://archive.stsci.edu/hlsp/jades 13 https://github.com/fengwusun/nircam_grism 3 The Astrophysical Journal, 997:61 (15pp), 2026 January 20 Lin et al. https://archive.stsci.edu/hlsp/jades https://github.com/fengwusun/nircam_grism catalog and line flux measurements will be publicly released along with F. Sun et al. (2025, in preparation). In this work, we only consider HAEs and AGNs with Hα luminosities greater than LHα > 1041.5 erg s−1. This is because the clustering analysis relies on luminosity function (LF) measurements, and the completeness correction for lower- luminosity bins in the Hα LFs is affected by large systematic uncertainties. For AGNs, this cut is applied to the total LHα to account for selection effects in the detectability of the Hα line. We further group galaxies with separations smaller than 10 physical kpc and 500 km s−1 into a system, assuming they are interacting and gravitationally bound. This separation corre- sponds to approximately 1.5 at z ≈ 4–6, consistent with the definition of a galaxy system applied to z ∼ 6 [O III] emitters (J. Matthee et al. 2023; A.-C. Eilers et al. 2024) and well within the defining separation for mergers (e.g., D. Puskás et al. 2025). This differs from the criteria set for the LF calculation in X. Lin et al. (2025), but it follows the same method used for the HAE clustering analysis in that paper. We finally obtain 782 systems at 3.9 < z < 6. Throughout this paper, we refer to these combined systems as galaxies for simplicity. 2.4. AGN Sample at 3.9 < z < 6 We selected AGNs from the parent HAE sample using three different sets of criteria: WFSS-based, V-shaped-based, and NIRSpec-based. All of these AGNs fall within the grism footprint. The selection criteria are outlined below. (1) WFSS-selected broad-line AGNs. With the F356W and F444W grism covering 3.1–5.0 μm, J. Zhang et al. (2025) identified 19 broad HAEs with compact morph- ology14 and FWHMs of broad Hα components exceed- ing 1000 km s−1. In our sample, seven of these AGNs were previously reported in J. Matthee et al. (2024). (2) NIRSpec-selected broad-line AGNs. R. Maiolino et al. (2023) identified high-redshift AGNs through prominent broad Hα emission lines in the NIRSpec R1000 grating spectra. In the GOODS-N field, they reported five objects at 4 < z < 6 with single robust broad components. We have excluded the tentatively detected broad-line candi- dates and the dual broad-line candidates from their sample. Four of the five AGNs fall within the WFSS footprint but do not overlap with the WFSS-selected AGNs. These four objects show bright Hα emission lines in the grism spectra, although the broad components are overwhelmed by the background noise because of the grism’s lower sensitivity and higher background level. (3) V-shaped continuum AGNs. We have searched for compact HAEs with V-shaped SEDs among the grism- selected HAEs. V-shaped SEDs, characterized by blue UV continuum slopes (βUV) and red optical continuum slopes (βopt), have been demonstrated as effective characteristics of high-redshift AGNs (e.g., J. E. Greene et al. 2024) especially when the broad Hα fluxes are below the grism detection limit. We do not require a broad Hα line detection for V-shaped-selected AGNs. First, we selected HAEs with a half-light radius 0 .2< in either the F444W or F356W filter. The robust spectroscopic redshifts (zspec) help us avoid contamination from brown dwarfs. We measured the UV continuum slope βUV by fitting a power-law model to the rest-frame range 1350–3000Å using the small-Kron-aperture KRON_S photometry (see JADES NIRCam data release; M. J. Rieke et al. 2023). The optical continuum slope βopt is derived from 4000–8000Å. We excluded bands that overlap with the Hβ + [O III], Hα lines, or the Balmer break around 3645Å. We require at least two filters spanning over 1000 Å in the rest frame to determine βopt. This strict requirement ensures that the measured slopes represent the pure continuum, free from contamination by strong emission lines and flux uncertainties. Among the 23 broad-line selected AGNs identified from WFSS and NIRSpec above, seven exhibit significant V-shaped SEDs with βUV < 0 and βopt > 0. Of the remaining 16 AGNs, seven have βopt < 0, while the other nine lack adequate band coverage to reliably determine βopt. In addition to the broad-line AGNs above, we identify five more HAEs with robust measurements of βUV < 0 and βopt > 0. Following the three selection criteria described above, we eventually selected a sample of 28 AGNs at z = 3.9–6 within the WFSS footprint for the parent HAE sample. Seven of them are also included in the photometrically selected sample in P. Rinaldi et al. (2024). The information of these 28 AGNs is summarized in Tables 1 and 2. Their spatial distribution is shown in Figure 1. All of these AGNs display Hα emission lines with LHα > 1041.5 erg s−1 in the grism spectra, satisfying the selection criteria for HAEs in the clustering analysis. 3. Diverse Environments of JWST-selected High- redshift AGNs In this section, we study the large-scale environments of AGNs at z = 3.9–6 based on the grism-selected HAEs. 3.1. Overdensity Field δ of High-redshift AGNs We calculate the overdensity field, δ, for each AGN based on the number density of HAEs within a volume of (15 h−1 cMpc)3, which equals the volumes defined in Y.-K. Chiang et al. (2013) and has been widely adopted in the literature (e.g., Y.-K. Chiang et al. 2014; O. Cucciati et al. 2014; Z. Cai et al. 2016; J. M. Helton et al. 2024; S. Lim et al. 2024). We center a cylinder with a radius of 8.5 h−1 cMpc and a length of 15 h−1 cMpc on the AGN, and calculate the galaxy density, n, within its volume by counting all enclosed galaxies. The WFSS footprint edge is taken into account. To calculate the mean galaxy density, n̄, we use the random galaxy catalog from the HAE autocorrelation analysis in X. Lin et al. (2025). The random galaxy catalog is designed to simulate a uniform distribution of galaxies without any clustering signal, while following the same selection function as the observed sample. The line fluxes of random galaxies are drawn from the Hα luminosity functions, and the completeness model is applied to account for selection effects. We count the random galaxies within the same volume and then scale the value to match the total HAE number density within the FOV. The overdensity field is computed as n n 1¯/= . The uncertainty of δ includes Poisson fluctuations of the observed galaxies and dispersion in 14 The criteria for compactness is defined based on circular aperture photometry 1.2F444W flux within r 0.2 F444W flux within r 0.1 = = , as justified in J. Zhang et al. (2025). 4 The Astrophysical Journal, 997:61 (15pp), 2026 January 20 Lin et al. the random galaxy distribution across different random galaxy realizations. The overdensity δ around AGNs is shown in Figure 2. The JWST AGNs are located in highly diverse large-scale environments. Their δ values range from −0.56, indicating underdense environments, to 10.56, representing an extreme overdensity. Among them, 10 AGNs (36% ± 9% assuming a binomial distribution) are in overdense regions with δ > 3, while the remaining 18 AGNs (64% ± 9%) are in average or underdense regions. We identify two extreme protoclusters at z = 4.41 and z = 5.19. The protocluster at z ≈ 4.41 contains five AGNs and 92 HAEs after applying the luminosity cut and grouping. These AGNs show overdensities ranging from δ = 6.85 to 10.51. Six AGNs at 5.16 < z < 5.29 are found in the filamentary structure at z = 5.19 (T. Herard-Demanche et al. 2023; F. Sun et al. 2024). The structure consists of three galaxy groups, with 93 galaxies at z = 5.19, 22 galaxies at z = 5.22, and 19 galaxies at z = 5.27. Among the six AGNs, two are associated with the z = 5.19 galaxy group, and two are associated with the z = 5.22 and z = 5.27 galaxy groups. These four exhibit δ = 5.17–10.56. The remaining two AGNs reside between these groups, with δ = 3.02 and 1.38, respectively. We show the 3D structures of both protoclusters in Figure 3. As a baseline for comparison, we also calculate the δ values for HAEs at similar redshifts. Among the 782 HAEs at 3.9 < z < 6, 232 (30% ± 2%) lie in regions with δ > 3. The fraction of galaxies in overdense regions is broadly consistent with that of the AGNs within 1σ. We conclude that, statistically, AGNs in our sample are not preferentially located in denser large-scale environments compared to star-forming galaxies. 3.2. Dependence of AGN Properties on Their Environments We explore the correlation between δ and the broad-line Hα luminosity (LHα,broad), the FWHMs of the broad Hα Table 1 Broad-line Selected AGN Sample in This Work ID R.A. Decl. zspec Selection Llog H ,broad FWHMHα,broad Mlog BH βUV βopt (erg s−1) (km s−1) (M⊙) 1087315 189.3336 62.2462 3.91 WFSS 42.09 ± 0.05 1514 ± 183 7.01 ± 0.11 −0.44 ± 1.02 −1.47 ± 0.13 1082263 189.2126 62.2274 3.98 WFSS 41.96 ± 0.07 1083 ± 205 6.65 ± 0.17 −1.70 ± 0.09 −0.20 ± 0.09 1089568 189.1518 62.2722 4.05 WFSS 42.22 ± 0.04 1461 ± 143 7.04 ± 0.09 −1.16 ± 0.43 −1.20 ± 0.10 1029154 189.1590 62.2602 4.17 WFSS 42.33 ± 0.04 2003 ± 225 7.38 ± 0.10 −1.56 ± 0.19 0.74 ± 0.17 1008411 189.2111 62.2503 4.41 WFSS 42.11 ± 0.11 3281 ± 757 7.72 ± 0.21 −0.74 ± 0.32 0.72 ± 0.12 1008671 189.1618 62.2511 4.41 WFSS 42.54 ± 0.02 2272 ± 95 7.59 ± 0.04 −1.53 ± 0.13 0.13 ± 0.11 1086855 189.2865 62.2381 4.41 WFSS 42.40 ± 0.06 1724 ± 271 7.28 ± 0.14 −1.01 ± 0.19 0.58 ± 0.12 1086784 189.3057 62.2369 4.41 WFSS 42.15 ± 0.07 3179 ± 504 7.71 ± 0.14 −0.13 ± 1.26 −0.10 ± 0.15 1033320 189.1258 62.2874 4.48 WFSS 42.04 ± 0.08 1951 ± 398 7.22 ± 0.19 −1.65 ± 0.12 −0.67 ± 0.26 1085355 189.0944 62.1990 4.88 WFSS 42.35 ± 0.03 1800 ± 158 7.29 ± 0.08 −1.76 ± 0.70 ⋯ 1090253 189.2855 62.2808 5.09 WFSS 42.41 ± 0.03 1455 ± 105 7.13 ± 0.07 −1.05 ± 0.75 ⋯ 1014406 189.0721 62.2734 5.15 WFSS 42.34 ± 0.07 3212 ± 639 7.81 ± 0.18 −1.61 ± 0.16 ⋯ 1034620 189.1598 62.2959 5.19 WFSS 42.68 ± 0.02 1077 ± 71 6.99 ± 0.06 −1.43 ± 0.06 ⋯ 1090549 189.2359 62.2855 5.19 WFSS 42.21 ± 0.07 1721 ± 307 7.19 ± 0.16 −1.88 ± 0.24 ⋯ 9994014 189.3001 62.2120 5.23 WFSS 42.61 ± 0.03 2084 ± 156 7.55 ± 0.07 2.04 0.93 0.70+ 2.00 0.82 1.12+ 1088832 189.3443 62.2634 5.24 WFSS 43.10 ± 0.02 2255 ± 91 7.85 ± 0.04 −1.27 ± 0.20 ⋯ 1013188 189.0571 62.2689 5.25 WFSS 42.33 ± 0.03 1957 ± 147 7.36 ± 0.07 −1.44 ± 0.29 ⋯ 1020514 189.1793 62.2925 5.36 WFSS 42.40 ± 0.03 1612 ± 140 7.22 ± 0.08 −1.91 ± 0.07 −0.32 ± 0.37 1087388 189.2810 62.2473 5.54 WFSS 43.52 ± 0.01 2964 ± 55 8.29 ± 0.02 −0.30 ± 0.19 ⋯ 1011836 189.2206 62.2637 4.41 NIRSpec 41.86 0.03 0.03+ 1451 105 98+ 7.13 ± 0.31 −0.94 ± 0.11 −2.14 ± 0.08 1020621 189.1225 62.2929 4.68 NIRSpec 42.00 0.03 0.03+ 1638 150 148+ 7.30 ± 0.31 −1.18 ± 0.17 0.88 ± 0.33 1001093 189.1797 62.2246 5.60 NIRSpec 41.86 0.03 0.04+ 1662 165 203+ 7.36 0.31 0.32+ −1.30 ± 0.30 0.12 ± 0.23 1061888 189.1680 62.2170 5.87 NIRSpec 42.15 0.02 0.03+ 1375 127 97+ 7.22 ± 0.31 −1.96 ± 0.17 ⋯ Note. ID refers to the source ID in the JADES DR3 photometric catalog (F. D’Eugenio et al. 2024), and zspec is the spectroscopic redshift measured from the NIRCam WFSS data. (ID 9994014 is not included in the JADES DR3 catalog because of a diffraction-spike mask, but is reported in J. Matthee et al. 2024.) SELECTION indicates the different selection criteria outlined in Section 2.4. The broad Hα luminosities (LHα,broad), FWHMs of the broad Hα emission lines (FWHMHα,broad), and black hole masses (MBH) for the WFSS-selected AGNs are taken from J. Zhang et al. (2025), while those for the NIRSpec-selected AGNs are from R. Maiolino et al. (2023). The UV (βUV) and optical (βopt) slopes are calculated using KRON_S photometry (Kron radius = 1.2) with at least two strong line-free bands spanning >1000 Å in the rest frame. Table 2 The V-shaped SED-selected AGNs in This Work, as a Complement to the Broad-line Sample in Table 1 ID R.A. Decl. zspec Selection βUV βopt 1055902 189.2348 62.2000 4.02 V-shaped SED −1.82 ± 0.19 0.05 ± 0.30 1029892 189.0928 62.2662 4.47 V-shaped SED −0.85 ± 0.23 0.14 ± 0.29 1066100 189.2272 62.2341 4.76 V-shaped SED −1.60 ± 0.27 0.79 ± 0.31 1081040 189.2816 62.2161 4.76 V-shaped SED −1.71 ± 0.20 1.30 ± 0.25 1020485 189.1131 62.2924 5.27 V-shaped SED −1.70 ± 0.23 0.60 ± 0.27 Note. ID refers to the source ID in the JADES DR3 photometric catalog, and zspec is the spectroscopic redshift measured from the NIRCam WFSS data. 5 The Astrophysical Journal, 997:61 (15pp), 2026 January 20 Lin et al. (FWHMHα,broad), and the BH mass (MBH). We use only the broad-line selected AGNs from the grism and NIRSpec (R. Maiolino et al. 2023; J. Zhang et al. 2025), for which these parameters can be measured. We note that precise MBH values are uncertain due to the unclear nature of these low- luminosity AGNs. We follow the common practice of deriving BH masses from LHα,broad and FWHMHα,broad to examine the potential correlation. In this work, MBH is calculated following A. E. Reines & M. Volonteri (2015), as a result of LHα,broad and FWHMHα,broad (M LFHWMBH H ,broad 2.06 H ,broad 0.47 ). We per- turb these parameters within their uncertainties to estimate the uncertainties in the correlation coefficient.15 The correlations between LHα,broad, FWHMHα,broad, and MBH with δ are shown in Figure 4. No clear correlation is found for LHα,broad, FWHMHα,broad, or MBH, as the Kendall’s τ analysis yields p > 0.05.16 We define the AGN fraction in this work as the proportion of AGNs that meet any of the three selection criteria (Section 2.4) among all grism-selected HAEs. The averages are 3.7% at 3.9 < z < 5, 4.0% at 5 < z < 6, and 3.9% for the combined range 3.9 < z < 6. We calculate the overdensity, δ, for all HAEs in the range 3.9 < z < 6 using the same method described in Section 3.1, and estimate the AGN fraction within each fixed δ range. Figure 5 shows the AGN fraction as a function of δ. The uncertainties are calculated assuming a binomial distribution for the AGN fraction. The Kendall’s τ analysis implies no significant dependence of AGN fraction on δ, with a correlation coefficient τ consistent with 0 and a p- value much greater than 0.05. We note that the AGN fractions shown here do not include corrections for selection functions or completeness, particularly for those selected through NIRSpec. For comparison, in Figure 5 we also show AGN fractions compiled from NIRSpec broad-line selections (Y. Harikane et al. 2023; R. Maiolino et al. 2023; I. Juodžbalis et al. 2025) and from grism-based broad-line selections (X. Lin et al. 2024; J. Matthee et al. 2024). Although the AGN fractions are luminosity-dependent and may be affected by cosmic variance, the absence of correlations in our results is less susceptible to these factors. A more accurate determina- tion of the AGN fraction will require future observations with higher sensitivity, more uniform selection criteria, and careful accounting for selection effects. The overdensity, δ, on 15 h−1 cMpc scales is primarily driven by the large-scale structure dominated by the linear two-halo terms (e.g., Y. Harikane et al. 2016; Y. Herrero Alonso et al. 2023). The lack of dependence of δ on MBH 108 731 5 108 226 3 105 590 2 108 956 8 102 915 4 101 183 6 100 841 1 100 867 1 108 685 5 108 678 4 102 989 2 103 332 0 102 062 1 106 610 0 108 104 0 108 535 5 109 025 3 101 440 6 103 462 0 109 054 9 999 401 4 108 883 2 101 318 8 102 048 5 102 051 4 108 738 8 100 109 3 106 188 8 0 5 10 ov er de ns ity ([1 5 h 1 c M pc ]3 ) 0. 85 ±0 .3 5 z= 3.9 1 1. 30 ±0 .3 4 z= 3.9 8 0. 86 ±0 .3 0 z= 4.0 2 1. 58 ±0 .4 1 z= 4.0 5 0. 85 ±0 .2 4 z= 4.1 7 7. 97 ±1 .0 2 z= 4.4 1 7. 68 ±0 .9 6 z= 4.4 1 6. 85 ±0 .9 0 z= 4.4 1 9. 99 ±1 .3 6 z= 4.4 1 10 .5 1± 1. 53 z= 4.4 1 0. 23 ±0 .0 9 z= 4.4 7 0. 06 ±0 .0 3 z= 4.4 8 -0 .5 6± 0. 40 z= 4.6 8 0. 13 ±0 .0 4 z= 4.7 6 0. 10 ±0 .0 4 z= 4.7 6 0. 92 ±0 .3 1 z= 4.8 8 0. 36 ±0 .1 6 z= 5.0 9 0. 48 ±0 .2 4 z= 5.1 5 6. 61 ±1 .3 7 z= 5.1 9 10 .5 6± 1. 64 z= 5.1 9 6. 80 ±1 .4 6 z= 5.2 3 3. 02 ±0 .9 7 z= 5.2 4 1. 38 ±0 .5 7 z= 5.2 5 5. 17 ±1 .3 1 z= 5.2 7 -0 .1 7± 0. 10 z= 5.3 6 0. 27 ±0 .1 2 z= 5.5 4 1. 76 ±0 .4 9 z= 5.6 0 -0 .4 8± 0. 34 z= 5.8 7 Figure 2. The overdensity fields δ over a volume of (15 h−1 cMpc)3 for AGNs at 3.9 < z < 6. The AGNs are ordered and color-coded by their redshift, with their IDs shown on the x-axis. The δ value is labeled at the top of each AGN. 505 RA (cMpc) 20 10 0 10 20 LOS distance (cMpc) 5 0 5 Dec (cM pc) AGN Protocluster Field galaxies 4.400 4.425 4.450 z sp ec 505 RA (cMpc) 20 10 0 10 20 30 40 50 LOS distance (cMpc) 5 0 5 Dec (cM pc) Group-1 Group-2 Group-3 Field galaxies AGN 5.2 5.3 zspec Figure 3. The 3D large-scale structure of the protocluster at z ≈ 4.41 (left) and z ≈ 5.19 (right), after the luminosity cut of 1041.5 erg s−1 and grouping within 500 pkpc and 500 km s−1. They host five and six AGNs, respectively. The coordinates of the z ≈ 4.41 structure are with respect to (R.A., decl., z) = (189.21355, 62.24861, 4.41), and the coordinates of the z ≈ 5.19 structure are with respect to (R.A., decl., z) = (189.20403, 62.23787, 5.195). All sources are color-coded by their redshifts, with the plus indicating AGNs, hexagons/squares indicating the protocluster members, and dots indicating field galaxies. 15 We use pymccorrelation (https://github.com/privong/ pymccorrelation) to implement the perturbation (P. A. Curran 2014; G. C. Privon et al. 2020). 16 The p-value denotes the probability of obtaining the current result if the correlation coefficient were zero (no correlation). If p is lower than 0.05, the correlation coefficient is statistically significant. 6 The Astrophysical Journal, 997:61 (15pp), 2026 January 20 Lin et al. https://github.com/privong/pymccorrelation https://github.com/privong/pymccorrelation implies significant scatter in the relation between large-scale environments and black hole (BH) mass in high-redshift AGNs. On the other hand, the bolometric luminosity and Eddington ratios can be derived using LHα,broad and FWHMHα,broad by applying the empirical relations for typical type-1 AGNs to these low-luminosity AGNs (J. E. Greene & L. C. Ho 2005; B. Trakhtenbrot et al. 2011), although this may not be accurate but is commonly adopted (e.g., R. Maiolino et al. 2023; X. Lin et al. 2024; J. Matthee et al. 2024). In this context, the lack of correlation of δ with respect to LHα,broad and FWHMHα,broad suggests that bolometric luminosity and Eddington ratios do not correlate with δ either. These results suggest that the large-scale environment on scales >10 cMpc does not significantly affect BH growth and AGN evolution on smaller, parsec scales. In contrast, J. Matthee et al. (2024b) found that overdensity within 1 cMpc is positively correlated with MBH, particularly when including measurements from 10 high-redshift AGNs and five quasars (A.-C. Eilers et al. 2024). X. Lin et al. (2024) also reported tentative evidence of small-scale clustering around one low-luminosity AGNs, with three neighboring galaxies located within 30 kpc. Our WFSS observations are not as deep as those in J. Matthee et al. (2024b), which were conducted in a lensing cluster field with longer exposures, limiting our ability to probe overdensities within 1 cMpc. The 1 cMpc scale is around the typical transition point between the dominance by nonlinear one-halo terms and linear two-halo terms for high-redshift galaxies (e.g., Y. Harikane et al. 2016; Y. Herrero Alonso et al. 2023). The differing environmental effects on MBH at scales <1 cMpc and >10 cMpc suggest that BH growth may be more strongly influenced by local nonlinear processes such as mergers. Alternatively, processes occurring on cosmological scales, such as the gas accretion from the large-scale cosmic web, have a weaker impact. These large-scale activities may require a longer time to come into effect and impact the small-scale BH activities. Future deep spectroscopic surveys are required to provide further insights by probing multiscale environments of a large AGN sample across a wide range of MBH. 4. The Clustering Analysis of High-redshift AGNs As illustrated in Section 3, there is significant variance in the environments of high-redshift AGNs. We therefore investigate their average population-level clustering properties in the context of large-scale structure of star-forming galaxies at the same cosmic epochs. For this clustering analysis, we focus exclusively on the 26 AGNs that meet the grism broad- line selection and V-shaped SED criteria. Two of the four NIRSpec-selected AGNs meet the V-shaped SED criteria and are therefore included in the clustering analysis. We do not include the remaining two NIRSpec-selected AGNs because of the limitation in quantifying their selection function via the Micro Shutter Array (MSA) design. The MSA targets must be pre-selected, resulting in a biased search. 4.1. The Projected Surface Density Excess We first compare the galaxy surface number density distributions centered on AGNs and HAEs. We define the Figure 4. The overdensity fields δ of AGNs vs. their broad Hα luminosities (left panel), FWHMs of their broad Hα lines (middle panel), and BH masses (right panel). The red dots represent the WFSS-selected broad-line AGNs, and the orange squares are the NIRSpec-selected broad-line AGNs. The Kendall’s τ analysis, with τ and p labeled on top of each panel, suggests no significant correlation between δ and the three parameters. 0 2 4 6 8 10 12 ([15h 1cMpc]3) 0 5 10 15 20 AG N fra cti on (% ) = 0.24+0.18 0.24 p = 0.38+0.37 0.24 Maiolino+23 Harikane+23 Matthee+24 Lin+24 Juod balis+25 This work Figure 5. AGN fraction among the star-forming galaxies vs. the overdensity field δ. The Kendall’s τ statistic and the associated probability p are labeled at the top left, revealing no significant correlation between AGN fraction and δ. The AGN fractions from literature studies are shown as green, blue, cyan, and orange shaded regions and green dashed–dotted lines for reference (Y. Harikane et al. 2023; R. Maiolino et al. 2023; X. Lin et al. 2024; J. Matthee et al. 2024; I. Juodžbalis et al. 2025). Note that these different AGN fractions are luminosity-dependent and subject to cosmic variance and selection effects. 7 The Astrophysical Journal, 997:61 (15pp), 2026 January 20 Lin et al. projected surface density excess within a radius of rp by r N r A r N A , 1p p p all all ( ) ( ) ( ) ( )< = < < where rp is the projected distance on the sky plane. N( 1 h−1 Mpc to that of the underlying dark matter field, assuming the transfer function model from CAMB (A. Lewis & A. Challinor 2011) and bias model of J. L. Tinker et al. (2010). We first obtain the bias for HAEs (bg) from the HAE autocorrelation function, fix it, and then fit the bias for AGNs (ba) following Equation (6). The results are reported in Table 3. Despite the large uncertainties due to the limited sample size, the halo masses of field AGNs inferred from UNIVERSEMACHINE (Mh) and from the bias measurements (Mh b, a) are consistent within 2σ. We adopt the UNIVERSEMA- CHINE estimates as fiducial. We note that, although available only for the field sample, adopting the bias-derived values does not alter the conclusions below. We compare the Mh of low-luminosity AGNs in our sample with luminous quasars across different cosmic epochs in Figure 8. For the clustering analysis of quasars or AGNs from large-area surveys (e.g., Sloan Digital Sky Survey, Subaru) with reported bias (N. P. Ross et al. 2009; Y. Shen et al. 2009; S. Eftekharzadeh et al. 2015; P. Laurent et al. 2017; J. D. Timlin et al. 2018; J. Arita et al. 2023, 2024), we convert their measured bias values into Mh using our adopted cosmological parameters and the halo bias models of J. L. Tinker et al. (2010) with the COLOSSUS17 package (B. Diemer 2018). We also present the median halo mass for quasars in A.-C. Eilers et al. (2024), which is derived by searching for analog systems to the quasar fields in UNIVERSEMACHINE. The bias in J. Arita et al. (2024) from the projected (angular) correlation analysis of 28 low- luminosity broad-line AGNs with photometrically selected galaxies corresponds to M Mlog 11.5 0.2 11.4h 0.3 0.2( ) ( )/ = ± + , based on the conversion in this study. These values are consistent with our measurements, accounting for sample variance and systematic uncertainties. In contrast, the host halo masses of luminous quasars at z = 0–6 typically exceed 1012M⊙, 1–2 dex higher than those of low-luminosity AGNs, while with little dependence on the quasar luminosity (e.g., Y. Shen et al. 2009). This implies that the differences between low-luminosity AGNs and UV-luminous quasars are not mainly caused by the obscuration from different geometries (H. Netzer 2015). These newly discovered AGNs are not simple counterparts of UV- bright quasars in dust-enshrouded environments. The low- luminosity AGNs discovered by JWST may be inherently distinct from luminous quasars. Instead, they reside in considerably less-massive dark matter halos. Given the comparable halo masses of AGNs and HAEs, along with the high AGN fraction of up to≳17% at LHα ≈ 1043 erg s−1 (X. Lin et al. 2024), low-luminosity AGNs might represent either a common stage in galaxy evolution or a distinct phase in the BH- galaxy coevolution as discussed in E. Pizzati et al. (2024a). To explain the characteristics of low-luminosity AGNs, such as their weak X-ray/radio emission and variability, super-Eddington accretion mode has been proposed in many studies (e.g., J. E. Greene et al. 2024; K. Inayoshi et al. 2024; E. Lambrides et al. 2024b; G. Mazzolari et al. 2024). Super- Eddington accretion is also proposed, if these AGNs have low duty cycles (≲1%), to reconcile their BH masses (E. Pizzati et al. 2024a). In this case, the predicted host halo masses would be Mh ∼ 1011M⊙. Interestingly, this predicted Mh value is consistent with our measurements and Mh derived from an independent simulation suite. Since low-luminosity AGNs and star-forming galaxies occupy halos of comparable mass, the AGN fraction (3.6% in our sample) provides a rough estimate of the duty cycle. This number is broadly consistent with E. Pizzati et al. (2024a). Multiwavelength observations are essential for better characterizing this population and its accretion mode, includ- ing their stellar component, radiation field, and dust content. 17 https://bdiemer.bitbucket.io/colossus/index.html 10 The Astrophysical Journal, 997:61 (15pp), 2026 January 20 Lin et al. https://bdiemer.bitbucket.io/colossus/index.html Large-volume surveys are required to probe the multiscale environments of BH growth across a broad range of BH masses. 5.2. Implication for the Host Galaxy Stellar Masses Assuming the same stellar-to-halo mass ratio distribution, the stellar masses of AGN host galaxies should be also comparable to those of HAEs in the same redshift range. The HAE stellar mass M�, derived using the UNIVERMACHINE simulations, has a median value of 108.6M⊙ for 3.9 < z < 5, and 108.4M⊙ for 5 < z < 6, with 1σ scattering of approximately 0.8 dex in both cases. We note that the M� distributions derived here are in good agreement with the M� values obtained from the SED fitting. We adopt the HAEs’ M� as a proxy for the stellar masses of AGN host galaxies and present the M�–MBH relation of AGNs in Figure 9. We adopt the MBH values for the grism-selected broad- line AGNs as representatives (Table 1), which range from 106.85M⊙ to 10 7.99M⊙ at 3.9 < z < 5 with a median of 107.32M⊙, and from 10 6.99M⊙ to 10 8.29M⊙ at 5 < z < 6 with a median of 107.46M⊙. Our results are consistent with some of the measurements for individual AGNs (R. Maiolino et al. 2023; K. N. Hainline et al. 2024), although the latter exhibit a wide range of M� with large uncertainties. On the MBH–M� diagram, these high-redshift AGNs have smaller M� compared to those of local AGNs with similarMBH. The observed highMBH/M� ratio cannot be fully attributed to observational bias (J. Li et al. 2025; Y. Sun et al. 2025). The MBH–M� relation for these AGNs is also offset from the theoretical prediction (K. Inayoshi & K. Ichikawa 2024), which was developed to model both high-luminosity quasars and low-luminosity AGNs in the early Universe. These overmassive BHs, with higher MBH/M� ratios than predicted values, highlight the need for next-generation models to better understand BH seeding mechanisms and the coevolution of BHs and galaxies. We also caution that the estimate of MBH is highly uncertain and debated. V. Rusakov et al. (2025) suggested that Balmer Luminous QSOs Ross+09 Shen+09 Eftekharzadeh+15 Laurent+17 Timlin+18 Arita+23 Eilers+24 High-z AGNs This work z=3.9-5 This work z=5-6 Arita+24 (projected) Arita+24 (angular) Luminous QSOs Ross+09 Shen+09 Eftekharzadeh+15 Laurent+17 Timlin+18 Arita+23 Eilers+24 High-z AGNs This work z=3.9-5 This work z=5-6 Arita+24 (projected) Arita+24 (angular) 0 2 4 6 z 1010 1011 1012 1013 M h [M ] Figure 8. The redshift evolution of dark matter halo masses for quasars and AGNs. For the bias measurements reported in the literature (N. P. Ross et al. 2009; Y. Shen et al. 2009; S. Eftekharzadeh et al. 2015; P. Laurent et al. 2017; J. D. Timlin et al. 2018; J. Arita et al. 2023, 2024), we convert the bias values using our adopted cosmological parameters and the halo bias models from J. L. Tinker et al. (2010). The Mh value from A.-C. Eilers et al. (2024) represents the median halo mass of quasars derived based on UNIVERSEMACHINE. We show twoMh values from J. Arita et al. (2024) in green, derived based on the projected cross-correlation functions of low-luminosity AGNs and photometrically selected galaxies, and their angular cross-correlation, respectively. 7 8 9 10 11 12 log(M /M ) 5 6 7 8 9 10 11 lo g( M BH /M ) High-z QSOs Ding+23 Yue+24 Stone+24 High-z AGNs Maiolino+23 Harikane+23 Local AGNs Kormendy&Ho 13 (ellip) Kormendy&Ho 13 (bulge) Reines&Volonteri 15 Zhuang&Ho+23 Inayoshi+24 high-z AGN model This work (3.9 < z < 5) This work (5 < z < 6) Figure 9. MBH–M� relation. The orange and purple hexagons are positioned at the median MBH measured from broad Hα lines, and median M� for HAEs at similar redshift. The error bars represent the 16%–84% range. The literature MBH and M� values of quasars, high-redshift AGNs, and local AGNs are compiled from J. Kormendy & L. C. Ho (2013), A. E. Reines & M. Volonteri (2015), X. Ding et al. (2023), Y. Harikane et al. (2023), R. Maiolino et al. (2023), M.-Y. Zhuang & L. C. Ho (2023), M. A. Stone et al. (2024), and M. Yue et al. (2024b). The model from K. Inayoshi & K. Ichikawa (2024) is shown as the dashed gray line. 11 The Astrophysical Journal, 997:61 (15pp), 2026 January 20 Lin et al. lines may be broadened by electron scattering in regions with high electron column densities. This effect could lead to an overestimation of MBH by 1–2 dex. In contrast, I. Juodžbalis et al. (2025) argued that this claim is untenable. Regardless, the estimates of Mh and the derived M� from Mh are independent of MBH. If MBH is overestimated, and the true mass is 105–107M⊙ as suggested by V. Rusakov et al. (2025), the MBH–M� relation would better match the model prediction. 6. Summary In this paper, we study the large-scale environments of low- luminosity AGNs at 3.9 < z < 6 in the GOODS-North field. By combining the JWST/NIRCam F356W grism from the CONGRESS program and the F444W grism from the FRESCO program, we identify 782 HAEs within this redshift range. From this dataset, we construct a sample of 28 low- luminosity AGNs selected using both the grism and NIRSpec spectra, along with spec-z-confirmed V-shaped SEDs. Our main conclusions are as follows: 1. Low-luminosity AGNs reside in a diverse range of large- scale environments. These AGNs are found in regions with overdensity fields δ within 15 cMpc spanning from low densities of δ = −0.56 to high densities of δ = 10.56. Notably, five AGNs are located in a protocluster at z = 4.41 and seven AGNs lie in filamentary structures at z ≈ 5.19. Among the 28 AGNs, 10 (36% ± 9%) are located in regions with overdensity δ > 3, a fraction consistent with that of HAEs in δ > 3 regions (30% ± 2%). This implies that AGNs do not preferen- tially reside in denser environments compared to HAEs. 2. No clear correlations are found between the overdensity field (δ; within 15 h−1 cMpc of AGNs) and their broad Hα luminosities, broad Hα FWHMs, or BH masses. The AGN fraction among star-forming galaxies also shows no dependence on δ. These results suggest that large- scale environments (>10 cMpc) may not significantly influence BH growth and AGN evolution on smaller (pc) scales. 3. The projected surface density excess of AGNs is consistent with that of HAEs, indicating that star- forming galaxies do not cluster more strongly around AGNs. This qualitatively indicates that the dark matter halos of AGNs have masses similar to those of star- forming galaxies. 4. The cross-correlations between AGNs and HAEs exhibit an amplitude comparable to that of the HAE autocorrela- tions at both 3.9 < z < 5 and 5 < z < 6. When considering AGNs and HAEs in the fields alone, the correlation length of the AGN-HAE cross-correlation is 4.26 0.61 0.70+ h−1 cMpc with a power-law index γ = 1.59 at 3.9 < z < 5, and 7.66 0.81 0.90+ h−1 cMpc with γ = 1.63 at 5 < z < 6. These values are consistent with those of the HAE autocorrelations within 1σ. It suggests that AGNs and HAEs share similar bias parameters (ba ≈ bg) and, consequently, reside in dark matter halos of compar- able mass. 5. Adopting the halo mass distribution for HAEs derived using the UNIVERSEMACHINE simulation (X. Lin et al. 2025), we find that low-luminosity AGNs are hosted by dark matter halos with masses of M Mlog h( )/ =11.0–11.2, with 1σ scattering of 0.3–0.4 dex. TheirMh values are 1–2 dex lower than those of luminous quasars at similar redshifts. The less- biased host dark matter halos suggest that low-luminosity AGNs likely represent a distinct evolutionary phase or AGN population. Interestingly, our Mh ≈ 1011M⊙ estimate is consistent with the low-duty cycle scenarios required for super-Eddington accretion, as suggested in simulations (E. Pizzati et al. 2024a). 6. Assuming the same stellar-to-halo mass ratio for AGNs and HAEs, the stellar masses (M�) of AGN host galaxies are M Mlog 8.2 8.4( ) –/ = with a typical 1σ scattering of 0.8 dex. The low-luminosity AGNs have overmassive BHs, showing higher M�/MBH ratios compared to local type-1 AGNs and theoretical predictions (K. Inayoshi & K. Ichikawa 2024). To better understand the nature of the JWST-selected low- luminosity AGNs, deep, wider spectroscopic surveys with large AGN samples across a wide range of BH masses and luminosity ranges are required. Deep spectroscopic surveys with large AGN samples, such as SAPPHIRES (GO 6434, PI Egami; F. Sun et al. 2025), will bring insights into the impact of local conditions on BH growth by small-scale clustering analysis. Wide grism surveys, like COSMOS-3D (GO 5893; PI Kakiichi) and NEXUS (GO 5105; PI Shen, Y. Shen et al. 2024), will provide better constraints on the halo masses across different luminosities, while also helping to mitigate cosmic variance. Acknowledgments We thank the anonymous referee for providing constructive comments. We thank Nickolas Kokron, Michael Strauss, and Yin Li for very helpful discussions on the clustering analysis. X.L. and X.F. acknowledge support from the NSF award AST- 2308258. F.W. acknowledges support from NSF award AST- 2513040. X.L. and Z.C. acknowledge support from the National Key R&D Program of China (grant No. 2023YFA1605600) and Tsinghua University Initiative Scientific Research Program (No. 20223080023). A.J.B. acknowledges funding from the “First- Galaxies” Advanced Grant from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement No. 789056). B.E.R. acknowledges support from the NIRCam Science Team contract to the University of Arizona, NAS5-02015, and JWST Program 3215. S.T. acknowledges support by the Royal Society Research grant G125142. C.N.A.W. acknowledges JWST/ NIRCam contract to the University of Arizona NAS5-02015. R. M. acknowledges support by the Science and Technology Facilities Council (STFC), by the ERC through Advanced grant 695671 “QUENCH,” and by the UKRI Frontier Research grant RISE and FALL. R.M. also acknowledges funding from a research professorship from the Royal Society. This work is based on observations made with the NASA/ ESA Hubble Space Telescope and NASA/ESA/CSA James Webb Space Telescope. The data were obtained from the Mikulski Archive for Space Telescopes at the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-03127 for JWST. These observations are associated with program Nos. 1181 (JADES), 1895 (FRESCO), and 3577 (CONGRESS). Support for program No. 3577 was provided by NASA through a grant from the Space Telescope Science Institute, which is operated by the Association of 12 The Astrophysical Journal, 997:61 (15pp), 2026 January 20 Lin et al. Universities for Research in Astronomy, Inc., under NASA contract NAS 5-03127. The authors acknowledge the FRESCO team for developing their observing program with a zero- exclusive-access period. Data Availability The JWST data presented in this article were obtained from the Mikulski Archive for Space Telescopes (MAST) at the Space Telescope Science Institute. The data of the FRESCO survey (The FRESCO Collaboration 2023) is available at doi:10.17909/gdyc-7g80; the data of the CONGRESS survey is available at doi:10.17909/6rfk-6s81; the data of the JADES survey (The JADES Collaboration 2023) is available at doi:10. 17909/8tdj-8n28. Appendix AGN sample We show the NIRCam images of our grism-selected AGN sample in Figure A1, NIRSpec-selected sample in Figure A2, and V-shaped-selected sample in Figure A3. For the V-shaped- selected sample, we present their rest-frame SEDs in Figure A4, which clearly exhibit bluer UV continuum slopes than the optical continuum slopes. 1087315 z=3.91 1082263 z=3.98 1089568 z=4.05 1029154 z=4.17 1008411 z=4.41 1008671 z=4.41 1086855 z=4.41 1086784 z=4.41 1033320 z=4.48 1085355 z=4.88 1090253 z=5.09 1014406 z=5.15 1034620 z=5.19 1090549 z=5.19 9994014 z=5.23 1088832 z=5.24 1013188 z=5.25 1020514 z=5.36 1087388 z=5.54 Figure A1. The 19 AGNs in the sample, identified through their broad Hα emission lines from JWST/NIRCam WFSS. For each AGN, we present a 2″ × 2″ thumbnail composed of F356W (or F444W for z > 5), F200W, and F115W images. For the sources 1085355, 1086784, 1087315, 1087388, 1088832, 1089568, 1090253, and 1090549, F210M images are used in place of F200W. 13 The Astrophysical Journal, 997:61 (15pp), 2026 January 20 Lin et al. https://doi.org/10.17909/gdyc-7g80 https://doi.org/10.17909/6rfk-6s81 https://doi.org/10.17909/8tdj-8n28 https://doi.org/10.17909/8tdj-8n28 ORCID iDs Xiaojing Linaa https://orcid.org/0000-0001-6052-4234 Xiaohui Fanaa https://orcid.org/0000-0003-3310-0131 Fengwu Sunaa https://orcid.org/0000-0002-4622-6617 Junyu Zhangaa https://orcid.org/0000-0002-1574-2045 Eiichi Egamiaa https://orcid.org/0000-0003-1344-9475 Jakob M. Heltonaa https://orcid.org/0000-0003-4337-6211 Feige Wangaa https://orcid.org/0000-0002-7633-431X Haowen Zhangaa https://orcid.org/0000-0002-4321-3538 Andrew J. Bunkeraa https://orcid.org/0000-0002-8651-9879 Zheng Caiaa https://orcid.org/0000-0001-8467-6478 Zhiyuan Jiaa https://orcid.org/0000-0001-7673-2257 Xiangyu Jinaa https://orcid.org/0000-0002-5768-738X Roberto Maiolinoaa https://orcid.org/0000-0002-4985-3819 Maria Anne Pudokaaa https://orcid.org/0000-0003- 4924-5941 Pierluigi Rinaldiaa https://orcid.org/0000-0002-5104-8245 Brant Robertsonaa https://orcid.org/0000-0002-4271-0364 Sandro Tacchellaaa https://orcid.org/0000-0002-8224-4505 Wei Leong Teeaa https://orcid.org/0000-0003-0747-1780 Yang Sunaa https://orcid.org/0000-0001-6561-9443 Christopher N. A. Willmeraa https://orcid.org/0000-0001- 9262-9997 Chris Willottaa https://orcid.org/0000-0002-4201-7367 Yongda Zhuaa https://orcid.org/0000-0003-3307-7525 References Akins, H. B., Casey, C. M., Lambrides, E., et al. 2025, ApJ, 991, 37 Arita, J., Kashikawa, N., Matsuoka, Y., et al. 2023, ApJ, 954, 210 Arita, J., Kashikawa, N., Onoue, M., et al. 2025, MNRAS, 536, 3677 Bañados, E., Venemans, B. P., Mazzucchelli, C., et al. 2018, Natur, 553, 473 Baggen, J. F. W., van Dokkum, P., Brammer, G., et al. 2024, ApJL, 977, L13 Behroozi, P., Wechsler, R. H., Hearin, A. P., & Conroy, C. 2019, MNRAS, 488, 3143 Brammer, G. B., van Dokkum, P. G., & Coppi, P. 2008, ApJ, 686, 1503 Cai, Z., Fan, X., Peirani, S., et al. 2016, ApJ, 833, 135 Cammelli, V., Monaco, P., Tan, J. C., et al. 2025, MNRAS, 536, 851 Casey, C. M., Akins, H. B., Kokorev, V., et al. 2024, ApJL, 975, L4 Chen, H., Eilers, A.-C., Bosman, S. E. I., et al. 2022, ApJ, 931, 29 Chiang, Y.-K., Overzier, R., & Gebhardt, K. 2013, ApJ, 779, 127 Chiang, Y.-K., Overzier, R., & Gebhardt, K. 2014, ApJL, 782, L3 Comastri, A., Gilli, R., Marconi, A., Risaliti, G., & Salvati, M. 2015, A&A, 574, L10 Cucciati, O., Zamorani, G., Lemaux, B. C., et al. 2014, A&A, 570, A16 1011836 z=4.41 1020621 z=4.68 1001093 z=5.60 1061888 z=5.87 Figure A2. Similar to Figure A1 but for the four AGNs in the sample identified through their broad Hα emission lines from JWST/NIRSpec Grating R1000 spectra (R. Maiolino et al. 2023). 1055902 z=4.02 1029892 z=4.47 1066100 z=4.76 1081040 z=4.76 1020485 z=5.27 Figure A3. Similar to Figure A1 but for the five AGNs in the sample identified through their V-shaped SEDs. 0.1 0.2 0.3 0.5 0.8 Rest Wavelength ( m) 100 101 102 Re st- fra m e f (n Jy ) 1055902 0.1 0.2 0.3 0.5 0.8 Rest Wavelength ( m) 1029892 0.1 0.2 0.3 0.5 0.8 Rest Wavelength ( m) 1066100 0.1 0.2 0.3 0.5 0.8 Rest Wavelength ( m) 1081040 0.1 0.2 0.3 0.5 0.8 Rest Wavelength ( m) 1020485 Figure A4. The rest-frame SEDs of the five V-shaped-selected AGNs. All of the photometry shown is measured with Kron radii of 1.2. The continuum photometry is represented by black squares, while the bands containing emission lines (Hβ, [O III] λλ4960,5008, and Hα) are marked as orange squares. The wavelengths of the emission lines are indicated by dashed orange vertical lines. The blue and red dashed lines represent the best-fit power-law models to the UV and optical continuum SEDs, respectively, with the corresponding shaded regions indicating the 1σ uncertainty. 14 The Astrophysical Journal, 997:61 (15pp), 2026 January 20 Lin et al. https://orcid.org/0000-0001-6052-4234 https://orcid.org/0000-0003-3310-0131 https://orcid.org/0000-0002-4622-6617 https://orcid.org/0000-0002-1574-2045 https://orcid.org/0000-0003-1344-9475 https://orcid.org/0000-0003-4337-6211 https://orcid.org/0000-0002-7633-431X https://orcid.org/0000-0002-4321-3538 https://orcid.org/0000-0002-8651-9879 https://orcid.org/0000-0001-8467-6478 https://orcid.org/0000-0001-7673-2257 https://orcid.org/0000-0002-5768-738X https://orcid.org/0000-0002-4985-3819 https://orcid.org/0000-0003-4924-5941 https://orcid.org/0000-0003-4924-5941 https://orcid.org/0000-0002-5104-8245 https://orcid.org/0000-0002-4271-0364 https://orcid.org/0000-0002-8224-4505 https://orcid.org/0000-0003-0747-1780 https://orcid.org/0000-0001-6561-9443 https://orcid.org/0000-0001-9262-9997 https://orcid.org/0000-0001-9262-9997 https://orcid.org/0000-0002-4201-7367 https://orcid.org/0000-0003-3307-7525 https://doi.org/10.3847/1538-4357/ade984 https://ui.adsabs.harvard.edu/abs/2025ApJ...991...37A/abstract https://doi.org/10.3847/1538-4357/ace43a https://ui.adsabs.harvard.edu/abs/2023ApJ...954..210A/abstract https://doi.org/10.1093/mnras/stae2765 https://ui.adsabs.harvard.edu/abs/2025MNRAS.536.3677A/abstract https://doi.org/10.1038/nature25180 https://ui.adsabs.harvard.edu/abs/2018Natur.553..473B/abstract https://doi.org/10.3847/2041-8213/ad90b8 https://ui.adsabs.harvard.edu/abs//2024ApJ...977L..13B/abstract https://doi.org/10.1093/mnras/stz1182 https://ui.adsabs.harvard.edu/abs/2019MNRAS.488.3143B/abstract https://ui.adsabs.harvard.edu/abs/2019MNRAS.488.3143B/abstract https://doi.org/10.1086/591786 https://ui.adsabs.harvard.edu/abs/2008ApJ...686.1503B/abstract https://doi.org/10.3847/1538-4357/833/2/135 https://ui.adsabs.harvard.edu/abs/2016ApJ...833..135C/abstract https://doi.org/10.1093/mnras/stae2663 https://ui.adsabs.harvard.edu/abs/2025MNRAS.536..851C/abstract https://doi.org/10.3847/2041-8213/ad7ba7 https://ui.adsabs.harvard.edu/abs/2024ApJ...975L...4C/abstract https://doi.org/10.3847/1538-4357/ac658d https://ui.adsabs.harvard.edu/abs/2022ApJ...931...29C/abstract https://doi.org/10.1088/0004-637X/779/2/127 https://ui.adsabs.harvard.edu/abs/2013ApJ...779..127C/abstract https://doi.org/10.1088/2041-8205/782/1/L3 https://ui.adsabs.harvard.edu/abs/2014ApJ...782L...3C/abstract https://doi.org/10.1051/0004-6361/201425496 https://ui.adsabs.harvard.edu/abs/2015A&A...574L..10C/abstract https://ui.adsabs.harvard.edu/abs/2015A&A...574L..10C/abstract https://doi.org/10.1051/0004-6361/201423811 https://ui.adsabs.harvard.edu/abs/2014A&A...570A..16C/abstract Curran, P. A. 2014, arXiv:1411.3816 Davies, F. B., Hennawi, J. F., & Eilers, A.-C. 2019, ApJL, 884, L19 Davies, F. B., Hennawi, J. F., & Eilers, A.-C. 2020, MNRAS, 493, 1330 Desjacques, V., Jeong, D., & Schmidt, F. 2018, PhR, 733, 1 D’Eugenio, F., Cameron, A. J., Scholtz, J., et al. 2025, ApJS, 277, 4 Diemer, B. 2018, ApJS, 239, 35 Ding, X., Onoue, M., Silverman, J. D., et al. 2023, Natur, 621, 51 Eftekharzadeh, S., Myers, A. D., White, M., et al. 2015, MNRAS, 453, 2779 Eilers, A.-C., Davies, F. B., Hennawi, J. F., et al. 2017, ApJ, 840, 24 Eilers, A.-C., Hennawi, J. F., Decarli, R., et al. 2020, ApJ, 900, 37 Eilers, A.-C., Mackenzie, R., Pizzati, E., et al. 2024, ApJ, 974, 275 Eisenstein, D. J., Willott, C., Alberts, S., et al. 2023, arXiv:2306.02465 Fan, X., Bañados, E., & Simcoe, R. A. 2023, ARA&A, 61, 373 Farina, E. P., Schindler, J.-T., Walter, F., et al. 2022, ApJ, 941, 106 Gardner, J. P., Mather, J. C., Abbott, R., et al. 2023, PASP, 135, 068001 Greene, J. E., & Ho, L. C. 2005, ApJ, 630, 122 Greene, J. E., Labbe, I., Goulding, A. D., et al. 2024, ApJ, 964, 39 Geach, J. E., Sobral, D., Hickox, R. C., et al. 2012, MNRAS, 426, 679 Greene, T. P., Chu, L., Egami, E., et al. 2016, SPIE, 9904, 99040E Hainline, K. N., Johnson, B. D., Robertson, B., et al. 2024, ApJ, 964, 71 Harikane, Y., Ouchi, M., Ono, Y., et al. 2016, ApJ, 821, 123 Harikane, Y., Zhang, Y., Nakajima, K., et al. 2023, ApJ, 959, 39 Helton, J. M., Sun, F., Woodrum, C., et al. 2024, ApJ, 974, 41 Hennawi, J. F., Kist, T., Davies, F. B., & Tamanas, J. 2025, MNRAS, 539, 2621 Hennawi, J. F., Strauss, M. A., Oguri, M., et al. 2006, AJ, 131, 1 Herard-Demanche, T., Bouwens, R. J., Oesch, P. A., et al. 2025, MNRAS, 537, 788 Herrero Alonso, Y., Miyaji, T., Wisotzki, L., et al. 2023, A&A, 671, A5 Horne, K. 1986, PASP, 98, 609 Illingworth, G. 2017, HST Proposal, id.15027. Cycle 25 Inayoshi, K., & Ichikawa, K. 2024, ApJL, 973, L49 Inayoshi, K., Kimura, S., & Noda, H. 2025, PASJ, 77, 811 Inayoshi, K., Nakatani, R., Toyouchi, D., et al. 2022, ApJ, 927, 237 Juodžbalis, I., Maiolino, R., Baker, W. M., et al. 2025, arXiv:2504.03551 Kokorev, V., Caputi, K. I., Greene, J. E., et al. 2024, ApJ, 968, 38 Kokubo, M., & Harikane, Y. 2025, ApJ, 995, 24 Kormendy, J., & Ho, L. C. 2013, ARA&A, 51, 511 Lambrides, E., Chiaberge, M., Long, A. S., et al. 2024a, ApJL, 961, L25 Lambrides, E., Garofali, K., Larson, R., et al. 2024b, arXiv:2409.13047 Landy, S. D., & Szalay, A. S. 1993, ApJ, 412, 64 Laurent, P., Eftekharzadeh, S., Le Goff, J.-M., et al. 2017, JCAP, 2017, 017 Leung, G. C. K., Finkelstein, S. L., Pérez-González, P. G., et al. 2025, ApJ, 992, 26 Lewis, A., & Challinor, A., 2011 CAMB: Code for Anisotropies in the Microwave Background, Astrophysics Source Code Library, ascl:1102.026 Li, J., Silverman, J. D., Shen, Y., et al. 2025, ApJ, 981, 19 Li, Z., Inayoshi, K., Chen, K., Ichikawa, K., & Ho, L. C. 2025, ApJ, 980, 36 Lim, S., Tacchella, S., Schaye, J., et al. 2024, MNRAS, 532, 4551 Lin, X., Egami, E., Sun, F., et al. 2025, arXiv:2504.08028 Lin, X., Wang, F., Fan, X., et al. 2024, ApJ, 974, 147 Lyu, J., Alberts, S., Rieke, G. H., et al. 2024, ApJ, 966, 229 Ma, Y., Greene, J. E., Setton, D. J., et al. 2025, ApJ, 981, 191 Maiolino, R., Risaliti, G., Signorini, M., et al. 2025, MNRAS, 538, 1921 Maiolino, R., Scholtz, J., Curtis-Lake, E., et al. 2024, A&A, 691, A145 Matthee, J., Mackenzie, R., Simcoe, R. A., et al. 2023, ApJ, 950, 67 Matthee, J., Naidu, R. P., Brammer, G., et al. 2024, ApJ, 963, 129 Matthee, J., Naidu, R. P., Kotiwale, G., et al. 2025, ApJ, 988, 246 Mazzolari, G., Gilli, R., Maiolino, R., et al. 2024, arXiv:2412.04224 Netzer, H. 2015, ARA&A, 53, 365 Ni, Y., Di Matteo, T., Gilli, R., et al. 2020, MNRAS, 495, 2135 Oesch, P. A., Brammer, G., Naidu, R. P., et al. 2023, MNRAS, 525, 2864 Pérez-González, P. G., Barro, G., Rieke, G. H., et al. 2024, ApJ, 968, 4 Pizzati, E., Hennawi, J. F., Schaye, J., et al. 2025, MNRAS, 539, 2910 Pizzati, E., Hennawi, J. F., Schaye, J., et al. 2024, MNRAS, 534, 3155 Privon, G. C., Ricci, C., Aalto, S., et al. 2020, ApJ, 893, 149 Pudoka, M., Wang, F., Fan, X., et al. 2024, ApJ, 968, 118 Puskás, D., Tacchella, S., Simmonds, C., et al. 2025, MNRAS, 540, 2146 Regan, J., & Volonteri, M. 2024, OJAp, 7, 72 Reines, A. E., & Volonteri, M. 2015, ApJ, 813, 82 Rieke, M. J., Robertson, B., Tacchella, S., et al. 2023, ApJS, 269, 16 Rinaldi, P., Bonaventura, N., Rieke, G. H., et al. 2025, ApJ, 992, 71 Robertson, B., Johnson, B. D., Tacchella, S., et al. 2024, ApJ, 970, 31 Ross, N. P., Shen, Y., Strauss, M. A., et al. 2009, ApJ, 697, 1634 Rusakov, V., Watson, D., Nikopoulos, G. P., et al. 2025, arXiv:2503.16595 Schindler, J.-T., Hennawi, J. F., Davies, F. B., et al. 2025, NatAs, 9, 1732 Shen, Y., Strauss, M. A., Ross, N. P., et al. 2009, ApJ, 697, 1656 Shen, Y., Zhuang, M.-Y., Li, J., et al. 2024, arXiv:2408.12713 Stone, M. A., Lyu, J., Rieke, G. H., Alberts, S., & Hainline, K. N. 2024, ApJ, 964, 90 Sun, F., Egami, E., Pirzkal, N., et al. 2022, ApJL, 936, L8 Sun, F., Egami, E., Pirzkal, N., et al. 2023, ApJ, 953, 53 Sun, F., Helton, J. M., Egami, E., et al. 2024, ApJ, 961, 69 Sun, F., Fudamoto, Y., Lin, X., et al. 2025, arXiv:2503.15587 Sun, Y., Rieke, G. H., Lyu, J., et al. 2025, arXiv:2503.03675 Tee, W. L., Fan, X., Wang, F., & Yang, J. 2025, ApJL, 983, L26 The FRESCO Collaboration 2023, FRESCO: MAST High Level Science Product, MAST High Level Science Product, Mikulski Archive for Space Telescopes (MAST), STScI/MAST, doi:10.17909/gdyc-7g80 The JADES Collaboration 2023, JWST Advanced Deep Extragalactic Survey (JADES): MAST High Level Science Product, MAST High Level Science Product, Mikulski Archive for Space Telescopes (MAST), STScI/MAST, doi:10.17909/8tdj-8n28 Timlin, J. D., Ross, N. P., Richards, G. T., et al. 2018, ApJ, 859, 20 Tinker, J. L., Robertson, B. E., Kravtsov, A. V., et al. 2010, ApJ, 724, 878 Trakhtenbrot, B., Netzer, H., Lira, P., & Shemmer, O. 2011, ApJ, 730, 7 Trinca, A., Valiante, R., Schneider, R., et al. 2024, arXiv:2412.14248 Wang, B., Leja, J., de Graaff, A., et al. 2024, ApJL, 969, L13 Wang, F., Yang, J., Fan, X., et al. 2021, ApJL, 907, L1 Williams, C. C., Alberts, S., Ji, Z., et al. 2024, ApJ, 968, 34 Yang, J., Wang, F., Fan, X., et al. 2020, ApJL, 897, L14 Yang, J., Wang, F., Fan, X., et al. 2021, ApJ, 923, 262 Yue, M., Eilers, A.-C., Ananna, T. T., et al. 2024a, ApJL, 974, L26 Yue, M., Eilers, A.-C., Simcoe, R. A., et al. 2024b, ApJ, 966, 176 Zhang, J., Egami, E., Sun, F., et al. 2025, arXiv:2505.02895 Zhang, Z., Jiang, L., Liu, W., & Ho, L. C. 2025, ApJ, 985, 119 Zhuang, M.-Y., & Ho, L. C. 2023, NatAs, 7, 1376 15 The Astrophysical Journal, 997:61 (15pp), 2026 January 20 Lin et al. https://arxiv.org/abs/1411.3816 https://doi.org/10.3847/2041-8213/ab42e3 https://ui.adsabs.harvard.edu/abs/2019ApJ...884L..19D/abstract https://doi.org/10.1093/mnras/stz3303 https://ui.adsabs.harvard.edu/abs/2020MNRAS.493.1330D/abstract https://doi.org/10.1016/j.physrep.2017.12.002 https://ui.adsabs.harvard.edu/abs/2018PhR...733....1D/abstract https://doi.org/10.3847/1538-4365/ada148 https://ui.adsabs.harvard.edu/abs/2025ApJS..277....4D/abstract https://doi.org/10.3847/1538-4365/aaee8c https://ui.adsabs.harvard.edu/abs/2018ApJS..239...35D/abstract https://doi.org/10.1038/s41586-023-06345-5 https://ui.adsabs.harvard.edu/abs/2023Natur.621...51D/abstract https://doi.org/10.1093/mnras/stv1763 https://ui.adsabs.harvard.edu/abs/2015MNRAS.453.2779E/abstract https://doi.org/10.3847/1538-4357/aa6c60 https://ui.adsabs.harvard.edu/abs/2017ApJ...840...24E/abstract https://doi.org/10.3847/1538-4357/aba52e https://ui.adsabs.harvard.edu/abs/2020ApJ...900...37E/abstract https://doi.org/10.3847/1538-4357/ad778b https://ui.adsabs.harvard.edu/abs/2024ApJ...974..275E/abstract https://arxiv.org/abs/2306.02465 https://doi.org/10.1146/annurev-astro-052920-102455 https://ui.adsabs.harvard.edu/abs/2023ARA&A..61..373F/abstract https://doi.org/10.3847/1538-4357/ac9626 https://ui.adsabs.harvard.edu/abs/2022ApJ...941..106F/abstract https://doi.org/10.1088/1538-3873/acd1b5 https://ui.adsabs.harvard.edu/abs/2023PASP..135f8001G/abstract https://doi.org/10.1086/431897 https://ui.adsabs.harvard.edu/abs/2005ApJ...630..122G/abstract https://doi.org/10.3847/1538-4357/ad1e5f https://ui.adsabs.harvard.edu/abs/2024ApJ...964...39G/abstract https://doi.org/10.1111/j.1365-2966.2012.21725.x https://ui.adsabs.harvard.edu/abs/2012MNRAS.426..679G/abstract https://doi.org/10.1117/12.2231347 https://ui.adsabs.harvard.edu/abs/2016SPIE.9904E..0EG/abstract https://doi.org/10.3847/1538-4357/ad1ee4 https://ui.adsabs.harvard.edu/abs/2024ApJ...964...71H/abstract https://doi.org/10.3847/0004-637X/821/2/123 https://ui.adsabs.harvard.edu/abs/2016ApJ...821..123H/abstract https://doi.org/10.3847/1538-4357/ad029e https://ui.adsabs.harvard.edu/abs/2023ApJ...959...39H/abstract https://doi.org/10.3847/1538-4357/ad6867 https://ui.adsabs.harvard.edu/abs/2024ApJ...974...41H/abstract https://doi.org/10.1093/mnras/staf643 https://ui.adsabs.harvard.edu/abs/2025MNRAS.539.2621H/abstract https://ui.adsabs.harvard.edu/abs/2025MNRAS.539.2621H/abstract https://doi.org/10.1086/498235 https://ui.adsabs.harvard.edu/abs/2006AJ....131....1H/abstract https://doi.org/10.1093/mnras/staf030 https://ui.adsabs.harvard.edu/abs/2025MNRAS.537..788H/abstract https://ui.adsabs.harvard.edu/abs/2025MNRAS.537..788H/abstract https://doi.org/10.1051/0004-6361/202244693 https://ui.adsabs.harvard.edu/abs/2023A&A...671A...5H/abstract https://doi.org/10.1086/131801 https://ui.adsabs.harvard.edu/abs/1986PASP...98..609H/abstract https://ui.adsabs.harvard.edu/abs/2017hst..prop15027I/abstract https://doi.org/10.3847/2041-8213/ad74e2 https://ui.adsabs.harvard.edu/abs/2024ApJ...973L..49I/abstract https://doi.org/10.1093/pasj/psaf050 https://ui.adsabs.harvard.edu/abs/2025PASJ...77..811I/abstract https://doi.org/10.3847/1538-4357/ac4751 https://ui.adsabs.harvard.edu/abs/2022ApJ...927..237I/abstract https://arxiv.org/abs/2504.03551 https://doi.org/10.3847/1538-4357/ad4265 https://ui.adsabs.harvard.edu/abs/2024ApJ...968...38K/abstract https://doi.org/10.3847/1538-4357/ae119e https://ui.adsabs.harvard.edu/abs/2025ApJ...995...24K/abstract https://doi.org/10.1146/annurev-astro-082708-101811 https://ui.adsabs.harvard.edu/abs/2013ARA&A..51..511K/abstract https://doi.org/10.3847/2041-8213/ad11ee https://ui.adsabs.harvard.edu/abs/2024ApJ...961L..25L/abstract https://arxiv.org/abs/2409.13047 https://doi.org/10.1086/172900 https://ui.adsabs.harvard.edu/abs/1993ApJ...412...64L/abstract https://doi.org/10.1088/1475-7516/2017/07/017 https://ui.adsabs.harvard.edu/abs/2017JCAP...07..017L/abstract https://doi.org/10.3847/1538-4357/adfcce https://ui.adsabs.harvard.edu/abs/2025ApJ...992...26L/abstract https://ui.adsabs.harvard.edu/abs/2025ApJ...992...26L/abstract http://www.ascl.net/1102.026 https://doi.org/10.3847/1538-4357/ada603 https://ui.adsabs.harvard.edu/abs/2025ApJ...981...19L/abstract https://doi.org/10.3847/1538-4357/ada5fb https://ui.adsabs.harvard.edu/abs/2025ApJ...980...36L/abstract https://doi.org/10.1093/mnras/stae1790 https://ui.adsabs.harvard.edu/abs/2024MNRAS.532.4551L/abstract https://arxiv.org/abs/2504.08028 https://doi.org/10.3847/1538-4357/ad6565 https://ui.adsabs.harvard.edu/abs/2024ApJ...974..147L/abstract https://doi.org/10.3847/1538-4357/ad3643 https://ui.adsabs.harvard.edu/abs/2024ApJ...966..229L/abstract https://doi.org/10.3847/1538-4357/ada613 https://ui.adsabs.harvard.edu/abs/2025ApJ...981..191M/abstract https://doi.org/10.1093/mnras/staf359 https://ui.adsabs.harvard.edu/abs/2025MNRAS.538.1921M/abstract https://doi.org/10.1051/0004-6361/202347640 https://ui.adsabs.harvard.edu/abs/2024A&A...691A.145M/abstract https://doi.org/10.3847/1538-4357/acc846 https://ui.adsabs.harvard.edu/abs/2023ApJ...950...67M/abstract https://doi.org/10.3847/1538-4357/ad2345 https://ui.adsabs.harvard.edu/abs/2024ApJ...963..129M/abstract https://doi.org/10.3847/1538-4357/ade886 https://ui.adsabs.harvard.edu/abs/2025ApJ...988..246M/abstract https://arxiv.org/abs/2412.04224 https://doi.org/10.1146/annurev-astro-082214-122302 https://ui.adsabs.harvard.edu/abs/2015ARA&A..53..365N/abstract https://doi.org/10.1093/mnras/staa1313 https://ui.adsabs.harvard.edu/abs/2020MNRAS.495.2135N/abstract https://doi.org/10.1093/mnras/stad2411 https://ui.adsabs.harvard.edu/abs/2023MNRAS.525.2864O/abstract https://doi.org/10.3847/1538-4357/ad38bb https://ui.adsabs.harvard.edu/abs/2024ApJ...968....4P/abstract https://doi.org/10.1093/mnras/staf660 https://ui.adsabs.harvard.edu/abs/2025MNRAS.539.2910P/abstract https://doi.org/10.1093/mnras/stae2307 https://ui.adsabs.harvard.edu/abs/2024MNRAS.534.3155P/abstract https://doi.org/10.3847/1538-4357/ab8015 https://ui.adsabs.harvard.edu/abs/2020ApJ...893..149P/abstract https://doi.org/10.3847/1538-4357/ad488a https://ui.adsabs.harvard.edu/abs/2024ApJ...968..118P/abstract https://doi.org/10.1093/mnras/staf813 https://ui.adsabs.harvard.edu/abs/2025MNRAS.540.2146P/abstract https://doi.org/10.33232/001c.123239 https://ui.adsabs.harvard.edu/abs/2024OJAp....7E..72R/abstract https://doi.org/10.1088/0004-637X/813/2/82 https://ui.adsabs.harvard.edu/abs/2015ApJ...813...82R/abstract https://doi.org/10.3847/1538-4365/acf44d https://ui.adsabs.harvard.edu/abs/2023ApJS..269...16R/abstract https://doi.org/10.3847/1538-4357/adfa10 https://ui.adsabs.harvard.edu/abs/2025ApJ...992...71R/abstract https://doi.org/10.3847/1538-4357/ad463d https://ui.adsabs.harvard.edu/abs/2024ApJ...970...31R/abstract https://doi.org/10.1088/0004-637X/697/2/1634 https://ui.adsabs.harvard.edu/abs/2009ApJ...697.1634R/abstract https://arxiv.org/abs/2503.16595 https://doi.org/10.1038/s41550-025-02660-1 https://ui.adsabs.harvard.edu/abs/2025NatAs...9.1732S/abstract https://doi.org/10.1088/0004-637X/697/2/1656 https://ui.adsabs.harvard.edu/abs/2009ApJ...697.1656S/abstract https://arxiv.org/abs/2408.12713 https://doi.org/10.3847/1538-4357/ad2a57 https://ui.adsabs.harvard.edu/abs/2024ApJ...964...90S/abstract https://ui.adsabs.harvard.edu/abs/2024ApJ...964...90S/abstract https://doi.org/10.3847/2041-8213/ac8938 https://ui.adsabs.harvard.edu/abs/2022ApJ...936L...8S/abstract https://doi.org/10.3847/1538-4357/acd53c https://ui.adsabs.harvard.edu/abs/2023ApJ...953...53S/abstract https://doi.org/10.3847/1538-4357/ad07e3 https://ui.adsabs.harvard.edu/abs/2024ApJ...961...69S/abstract http://arXiv.org/abs/2503.15587 https://arxiv.org/abs/2503.03675 https://doi.org/10.3847/2041-8213/adc5e3 https://ui.adsabs.harvard.edu/abs/2025ApJ...983L..26T/abstract https://doi.org/10.17909/gdyc-7g80 https://doi.org/10.17909/8tdj-8n28 https://doi.org/10.3847/1538-4357/aab9ac https://ui.adsabs.harvard.edu/abs/2018ApJ...859...20T/abstract https://doi.org/10.1088/0004-637X/724/2/878 https://ui.adsabs.harvard.edu/abs/2010ApJ...724..878T/abstract https://ui.adsabs.harvard.edu/abs/2010ApJ...724..878T/abstract https://doi.org/10.1088/0004-637X/730/1/7 https://ui.adsabs.harvard.edu/abs/2011ApJ...730....7T/abstract https://arxiv.org/abs/2412.14248 https://doi.org/10.3847/2041-8213/ad55f7 https://ui.adsabs.harvard.edu/abs//2024ApJ...969L..13W/abstract https://doi.org/10.3847/2041-8213/abd8c6 https://ui.adsabs.harvard.edu/abs/2021ApJ...907L...1W/abstract https://doi.org/10.3847/1538-4357/ad3f17 https://ui.adsabs.harvard.edu/abs/2024ApJ...968...34W/abstract https://doi.org/10.3847/2041-8213/ab9c26 https://ui.adsabs.harvard.edu/abs/2020ApJ...897L..14Y/abstract https://doi.org/10.3847/1538-4357/ac2b32 https://ui.adsabs.harvard.edu/abs/2021ApJ...923..262Y/abstract https://doi.org/10.3847/2041-8213/ad7eba https://ui.adsabs.harvard.edu/abs/2024ApJ...974L..26Y/abstract https://doi.org/10.3847/1538-4357/ad3914 https://ui.adsabs.harvard.edu/abs/2024ApJ...966..176Y/abstract http://arXiv.org/abs/2505.02895 https://doi.org/10.3847/1538-4357/adcb3e https://ui.adsabs.harvard.edu/abs/2025ApJ...985..119Z/abstract https://doi.org/10.1038/s41550-023-02051-4 https://ui.adsabs.harvard.edu/abs/2023NatAs...7.1376Z/abstract 1. Introduction 2. Data and Sample 2.1. Imaging and Photometric Catalog 2.2. JWST/NIRCam Grism Spectroscopy 2.3. Hα Emitters at 3.9 < z < 6 2.4. AGN Sample at 3.9 < z < 6 3. Diverse Environments of JWST-selected High-redshift AGNs 3.1. Overdensity Field δ of High-redshift AGNs 3.2. Dependence of AGN Properties on Their Environments 4. The Clustering Analysis of High-redshift AGNs 4.1. The Projected Surface Density Excess 4.2. The Cross-correlation between Hα Emitters and AGNs 5. Discussion 5.1. Implications for the AGN Host Dark Matter Halo Masses 5.2. Implication for the Host Galaxy Stellar Masses 6. Summary Data Availability Appendix. AGN sample References