We study the relation between stellar mass (M *) and star formation rate (SFR) for star-forming galaxies over approximately five decades in stellar mass ( 5.5 ≲ log 10 ( M * / M ⊙ ) ≲ 10.5 ) at z ≈ 3–6.5. This unprecedented coverage has been possible thanks to the joint analysis of blank non-lensed fields (COSMOS/SMUVS) and cluster lensing fields (Hubble Frontier Fields) that allow us to reach very low stellar masses. Previous works have revealed the existence of a clear bimodality in the SFR–M * plane with a star formation Main Sequence and a starburst cloud at z ≈ 4–5. Here we show that this bimodality extends to all star-forming galaxies and is valid in the whole redshift range z ≈ 3–6.5. We find that starbursts constitute at least ≈20% of all star-forming galaxies with M * ≳ 109 M ⊙ at these redshifts and reach a peak of 40% at z = 4–5. More importantly, 60%–90% of the total SFR budget at these redshifts is contained in starburst galaxies, indicating that the starburst mode of star formation is dominant at high redshifts. Almost all the low stellar mass starbursts with log 10 ( M * / M ⊙ ) ≲ 8.5 have ages comparable to the typical timescales of a starburst event, suggesting that these galaxies are being caught in the process of formation. Interestingly, galaxy formation models fail to predict the starburst/main-sequence bimodality and starbursts overall, suggesting that the starburst phenomenon may be driven by physical processes occurring at smaller scales than those probed by these models.

We present the ultraviolet luminosity function and an estimate of the cosmic star formation rate density at 8 < z < 13 derived from deep NIRCam observations taken in parallel with the MIRI Deep Survey of the Hubble Ultra Deep Field (HUDF), NIRCam covering the parallel field 2. Our deep (40 hr) NIRCam observations reach an F277W magnitude of 30.8 (5σ), more than 2 mag deeper than JWST public data sets already analyzed to find high-redshift galaxies. We select a sample of 44 z > 8 galaxy candidates based on their dropout nature in the F115W and/or F150W filters, a high probability for their photometric redshifts, estimated with three different codes, being at z > 8, good fits based on χ 2 calculations, and predominant solutions compared to z < 8 alternatives. We find mild evolution in the luminosity function from z ∼ 13 to z ∼ 8, i.e., only a small increase in the average number density of ∼0.2 dex, while the faint-end slope and absolute magnitude of the knee remain approximately constant, with values α = − 2.2 ± 0.1, and M * = − 20.8 ± 0.2 mag. Comparing our results with the predictions of state-of-the-art galaxy evolution models, we find two main results: (1) a slower increase with time in the cosmic star formation rate density compared to a steeper rise predicted by models; (2) nearly a factor of 10 higher star formation activity concentrated in scales around 2 kpc in galaxies with stellar masses ∼108 M ⊙ during the first 350 Myr of the universe, z ∼ 12, with models matching better the luminosity density observational estimations ∼150 Myr later, by z ∼ 9.

Mid-infrared (mid-IR) observations are powerful in identifying heavily obscured active galactic nuclei (AGN) that have weak emission in other wavelengths. Data from the Mid-Infrared Instrument (MIRI) on board the James Webb Space Telescope provides an excellent opportunity to perform such studies. We take advantage of the MIRI imaging data from the Cosmic Evolution Early Release Science Survey to investigate the AGN population in the distant universe. We estimate the source properties of MIRI-selected objects by utilizing spectral energy distribution (SED) modeling, and classify them into star-forming galaxies (SFs), SF-AGN mixed objects, and AGN. The source numbers of these types are 433, 102, and 25, respectively, from four MIRI pointings covering ∼9 arcmin2. The sample spans a redshift range of ≈0–5. We derive the median SEDs for all three source types, respectively, and publicly release them. The median MIRI SED of AGN is similar to the typical SEDs of hot dust-obscured galaxies and Seyfert 2s, for which the mid-IR SEDs are dominated by emission from AGN-heated hot dust. Based on our SED-fit results, we estimate the black hole accretion density (BHAD; i.e., total BH growth rate per comoving volume) as a function of redshift. At z < 3, the resulting BHAD agrees with the X-ray measurements in general. At z > 3, we identify a total of 27 AGN and SF-AGN mixed objects, leading to that our high-z BHAD is substantially higher than the X-ray results (∼0.5 dex at z ≈ 3–5). This difference indicates MIRI can identify a large population of heavily obscured AGN missed by X-ray surveys at high redshifts.

We make use of JWST medium-band and broadband NIRCam imaging, along with ultradeep MIRI 5.6 μm imaging, in the Hubble eXtreme Deep Field to identify prominent line emitters at z ≃ 7–8. Out of a total of 58 galaxies at z ≃ 7–8, we find 18 robust candidates (≃31%) for (Hβ + [O iii]) emitters, based on their enhanced fluxes in the F430M and F444W filters, with EW0(Hβ +[O iii]) ≃87–2100 Å. Among these emitters, 16 lie in the MIRI coverage area and 12 exhibit a clear flux excess at 5.6 μm, indicating the simultaneous presence of a prominent Hα emission line with EW0(Hα) ≃200–3000 Å. This is the first time that Hα emission can be detected in individual galaxies at z > 7. The Hα line, when present, allows us to separate the contributions of Hβ and [O iii] to the (Hβ +[O iii]) complex and derive Hα-based star formation rates (SFRs). We find that in most cases [O iii]/Hβ > 1. Instead, two galaxies have [O iii]/Hβ < 1, indicating that the NIRCam flux excess is mainly driven by Hβ. Most prominent line emitters are very young starbursts or galaxies on their way to/from the starburst cloud. They make for a cosmic SFR density log 10 ( ρ SFR H α / ( M ⊙ yr − 1 Mpc − 3 ) ) ≃ − 2.35 , which is about a quarter of the total value ( log 10 ( ρ SFR tot / ( M ⊙ yr − 1 Mpc − 3 ) ) ≃ − 1.76 ) at z ≃ 7–8. Therefore, the strong Hα emitters likely had a significant role in reionization.

Using the novel James Webb Space Telescope (JWST)/NIRCam observations in the A2744 field, we present a first spatially resolved overview of a Hubble Space Telescope (HST)-dark galaxy, spectroscopically confirmed at z = 2.58 with magnification μ ≈ 1.9. While being largely invisible at ∼1 μm with NIRCam, except for sparse clumpy substructures, the object is well detected and resolved in the long-wavelength bands with a spiral shape clearly visible in F277W. By combining ancillary Atacama Large Millimeter/submillimeter Array (ALMA) and Herschel data, we infer that this object is an edge-on dusty spiral with an intrinsic stellar mass log (M */M ⊙) ∼ 11.3 and a dust-obscured star formation rate ∼300 M ⊙ yr−1. A massive quiescent galaxy (log (M */M ⊙) ∼ 10.8) with tidal features lies 2.″0 away (r ∼ 9 kpc), at a consistent redshift as inferred by JWST photometry, indicating a potential major merger. The dusty spiral lies on the main sequence of star formation, and shows high dust attenuation in the optical (3 < A V < 4.5). In the far-infrared, its integrated dust spectral energy distribution is optically thick up to λ 0 ∼ 500 μm, further supporting the extremely dusty nature. Spatially resolved analysis of the HST-dark galaxy reveals a largely uniform A V ∼ 4 area spanning ∼57 kpc2, which spatially matches to the ALMA 1 mm continuum emission. Accounting for the surface brightness dimming and the depths of current JWST surveys, unlensed analogs of the HST-dark galaxy at z > 4 would be only detectable in F356W and F444W in an UNCOVER-like survey, and become totally JWST-dark at z ∼ 6. This suggests that detecting highly attenuated galaxies in the Epoch of Reionization might be a challenging task for JWST.