There is considerable debate about the formation of massive stars, including whether a high-mass star must always form with a population of low-mass stars, or if it can also form in isolation. Massive stars found in the field are often considered to be runaways from star clusters or OB associations. However, there is evidence in the Milky Way and the Small Magellanic Cloud of high-mass stars that appear to be isolated in the field and they cannot be related to any known star cluster or OB association. Studies of more distant galaxies have been lacking so far. We identified massive star candidates that appear isolated in the field of the nearby spiral galaxy NGC 4242 (at a distance of 5.3 Mpc) to explore how many candidates for isolated star formation we find in a galaxy outside the Local Group. We identified 234 massive (M_ ini and young (łeq 10 Myr) field stars in NGC 4242 using the Hubble Space Telescope Solar Blind Channel of the Advanced Camera for Surveys, the UVIS channel of the Wide Field Camera 3 from the Galaxy UV Legacy Project (GULP), and optical data from the Legacy ExtraGalactic UV Survey (LEGUS). We investigated the surroundings of our targets within the range of projected distances expected for runaway stars, 74 pc and 204 pc. Within the threshold radii, 9.8% and 34.6% of our targets have no young star clusters, OB associations, or massive stars. This causes them to appear isolated. This fraction reduces to 3.2%-11.5% for the total number of massive stars expected from the observed UV star formation rate. Our results show that there is a small population of young and massive potentially isolated field stars in NGC 4242.

One of the most remarkable and unexpected results of the James Webb Space Telescope is the discovery of a population of compact red galaxies: the so-called little red dots (LRDs). The existence of these galaxies raises many questions, including that of their nature and origin, as well as that of their evolution. Indeed, these compact red sources exhibit a pronounced decline in number density by nearly two orders of magnitude from ( z = 6 ) to ( z = 3 ). In this paper, we investigate the possible evolution of this galaxy population at a lower redshift. To this end, we have identified a sample of candidates in the CEERS images that could represent the descendants of LRDs by assuming a single evolutionary path: the development of a blue star-forming outskirt while retaining a inner red core. Our color–magnitude selection identifies red galaxies as red as LRDs at ( z < 4 ), defined by a compact, red, inner region and blue outskirts. The red core is associated with the LRD population, while the blue periphery traces recently formed young stars. Morphological properties were derived by fitting single Sérsic profiles, while other physical quantities were obtained through spectral energy distribution (SED) fitting, assuming a stellar-only model for both the inner region and the outskirts. The selected galaxies are likely "post-LRDs" galaxies, showing similar properties to LRDs under a stellar-only model: stellar masses of M_* ≈ 10^ 10 , M_⊙, central densities Σ_* ≈ 11 , M_⊙ , kpc ^ -2 , similar rest-frame red colors, and a ∼1 kpc offset below the size–mass relation. Their number density at z=3 ± 0.5 ($ 10^ -4.15 , Mpc ^ -3 $) matches that of LRDs at 5 < z < 7, supporting an evolutionary connection. We find a strong redshift-dependent increase in both outskirts' mass fraction and galaxy size, from ∼250 pc at z=5 to ∼600 pc at z=3, indicating overall stellar growth. Meanwhile, the core remains as red and as massive, but the characteristic V shaped SED fades as the extended star-forming envelope becomes dominant. These findings support an evolutionary scenario in which LRDs gradually acquire an extended stellar component over cosmic time by cold accretion. This growth affects the initial LRD state in two key ways: the physical size increases and the characteristic V shaped SED in the core becomes less distinct and disappears. As a result, the original selection criteria based on both of them can no longer identify this population as it evolves, providing an explanation for their observed decline in number density.

Photometric variability is a defining characteristic of young stellar objects (YSOs) that can be traced back to a range of physical processes that occur at different stages in the formation and early evolution of young stars. The Gaia third Data Release (GDR3) has provided an unprecedented dataset of photometric time series, including 79,375 light curves for sources classified as YSO candidates. Through its all-sky coverage, Gaia provides a unique opportunity for large-scale studies of YSO variability. Our goal was to characterise the GDR3 sample of YSO variables to better identify the recurrence of YSO variability modes (caused by accretion, extinction, rotation modulation, etc.). We made a pilot study of the applicability of the asymmetry (M) and periodicity (Q) variability metrics to characterise YSO variability with Gaia light curves. By adapting the Q-M metrics for sparse and long-term light curves, we sought to bridge the gap between low- and high-cadence survey insights on YSO variability. We adapted the Q-M method for Gaia. Through a refined sample selection, we identified sources with an appropriate sampling for the Q-M method. We used the generalised Lomb Scargle periodogram and structure functions to derive characteristic variability timescales. We successfully derived Q-M indices for 23,000 sources in the GDR3 YSO sample. These variables were then classified into eight variability morphological classes. We linked the morphological classes with physical mechanisms using Hα as a proxy of accretion and α_ indices to gauge whether circumstellar material was present. We demonstrate that the Q-M metrics can be successfully applied to study the sparse time series of Gaia. We applied it successfully to distinguish between the various variability modes of YSOs. While our results are generally consistent with previous high-cadence short-term studies, the long GDR3 time span yields a larger variety of variability mechanisms.

Evolution of the Accretion Rate of Young Intermediate-mass Stars: Implications for Disk Evolution and Planet Formation

We update the status of the spectropolarimetric campaign dedicated to characterise the magnetic field properties of a sample of known exoplanet-hosting stars included in the current target list of the Ariel mission. The main aims are to inform observing strategies and subsequent analysis of the data of the Ariel mission, and to provide background information on the magnetic properties of the target and their variability on timescales of at least a few years. We analysed spectropolarimetric data collected for 15 G-M type stars with Neo-Narval, HARPSpol, and SPIRou to assess the detectability of the large-scale magnetic field. For three stars we reconstructed the magnetic field topology and its temporal evolution via Zeeman-Doppler imaging (ZDI). Such reconstructions were then used to perform three-dimensional magnetohydrodynamical simulations of the stellar wind and environment impinging on the hosted exoplanets. We detected the magnetic field of six stars. Of these, we performed ZDI reconstructions for the first time of TOI-1860 and DS Tuc A, and for the second time of HD 63433, providing temporal information of its large-scale magnetic field. Consistently with previous results on young (̊m∼ 50-100 Myr) solar-like stars, the large-scale magnetic field is moderately strong (30-60,G on average) and complex, with a significant fraction of magnetic energy in the toroidal component and high-order poloidal components. From the simulations of the stellar wind, we found the orbit of TOI-1860 b to be almost completely sub-Alfvénic, the orbits of DS Tuc A b and HD 63433 d to be trans-Alfvénic, and the orbits of HD 63433 b and c to be super-Alfvénic. We obtained marginal detections of the magnetic field for TOI-836 and TOI-2076, and detections for TOI-1136, but the number of observations is not sufficient for magnetic mapping. A magnetic star-planet connection can occur for most of TOI-1860 b's orbit. This can happen more sporadically for DS Tuc A b and HD 63433 c given the lower fraction of their orbit in the sub-Alfvénic regime. The orbit of HD 63433 c is nevertheless more sub-Alfvénic than previously simulated owing to the temporal evolution of the stellar magnetic field. For HD 63433 b and c, we expect the formation of a bow shock between the stellar wind and the planet despite the evolution of the stellar magnetic field.

Abstract We have carried out a photometric time domain study of 188 intermediate-mass young stars observed in Full Frame Image mode with the Transiting Exoplanet Survey Satellite (TESS) satellite over the first 3.3 yr of its mission. The majority of these targets are classified as Herbig Ae/Be stars (HAeBes). All were monitored at optical wavelengths for at least one 27 day TESS sector, with many having multiple sectors of data. From a custom aperture photometry pipeline, we produced light curves and analyzed the variability therein, as a function of stellar and circumstellar properties. Based on visual and statistical analysis, we find that ∼95% of HAeBes are variable on timescales of 10 minutes to one month, with the most common light-curve morphology being stochastic. Approximately 15% of the set display quasiperiodic variability. In comparison to sets of low-mass T Tauri stars monitored with optical space telescopes, the Herbig Ae/Be stars display a much lower incidence of “dipper” behaviors (quasiperiodic or aperiodic fading events), as well as periodic modulations. As posited by previous work, we conclude that magnetic starspots are rare on HAeBes, and that the inner circumstellar dust rims of these objects lie at substantially larger radii than for low-mass young stars. Beyond these differences, the accretion dynamics of young stars less than ∼7 M ⊙ appear to be largely consistent based on their time domain properties from data streams of up to three months’ duration. We do, however, find tentative evidence for a change in variability amplitude above this mass boundary, particularly for quasiperiodic behavior.

Abstract Young stars form in clusters within molecular clouds, but older stars are evenly distributed across the galactic disk, necessitating an explanation for cluster dissolution. We analytically study tidal forces from cold molecular clouds as a key mechanism for accelerated cluster disruption. Cloud tides, caused by the gravitational pull of the parent cloud along the radial direction, arise from the spatial gradient of gravitational acceleration and drive cluster disruption. This mechanism activates after gas expulsion and remains effective until the cloud is disrupted by stellar feedback or the cluster moves away. Cloud tides act on gas-deprived clusters, causing exponential expansion on a tidal timescale of t tidal , ext = 3 / ( 8 π G ρ mean ) , where ρ mean is the cloud’s density at the cluster’s location. With a duration of a few Myr, cloud tides can lead to a 10 times increase of the cluster size, producing bar-like elongated stellar aggregations resembling Gaia strings. These results establish cloud tides as a potentially important mechanism for star cluster disruption.

Abstract We re-evaluate the star formation history of the nearby Scorpius-Centaurus (Sco-Cen) OB Association with a comprehensive analysis of Gaia XP spectra of more than 7800 potential members. New spectral classifications are obtained by fitting individual XP spectra with templates derived from empirical spectra of young stars. Combining these spectral classifications in this work and in the literature with archival photometry leads to estimates of V -band extinction and stellar luminosities for a total of 8846 sources. Employing SPOTS models with spot coverages of 0.34 and 0.51 for K- and M-type stars harmonizes age estimates between K/M-type and F/G-type stars, with ages older than are obtained for low-mass stars from standard evolutionary models. These older ages lead to a disk lifetime that is approximately two times longer than reported in previous literature. Our re-evaluation of the star formation history with these revised age estimates uncovers evidence of underlying substructures within the Sco-Cen complex.

We investigate the impacts of the evolution of dust mass and grain size distribution on the evolution of global attenuation curves, with a focus on the optical-ultraviolet (UV) slope and the $2175$ Å bump, within a Milky Way-like (MW-like) galaxy simulation. In addition, we discuss the contributions of the star-dust geometry, scattering, and dust properties to the attenuation curves. We performed the post-processing dust radiative transfer using the SKIRT code based on a MW-like galaxy simulation. The hydrodynamic simulation was carried out with the GADGET4-OSAKA code, which models the evolution of grain size distributions. For lower inclination angles (i.e., closer to face-on), the attenuation curve flattens over time up to t = 1 Gyr and becomes progressively steeper. The steeper slope of the attenuation curve is caused by the interplay between scattering and the dust disk becoming more extended over time (i.e., changes in the star-dust geometry). At higher inclination angles, the effect of scattering is suppressed and the attenuation curves steepen slightly over time due to the formation of small grains and the bias of observed UV emission toward old stars. The $2175$ Å bump becomes stronger on a timescale of ∼250 Myr due to the formation of small carbonaceous grains. However, the bump strength is affected not only by the abundance of small grains, but also by star-dust geometry. At higher A_V, or at higher inclination angles, the bump strengths become weaker. These results may help interpret flatter attenuation curves and less prominent bumps in high-redshift galaxies. Furthermore, we find that variations in the star-dust geometry alter the amount of scattered photons escaping the galaxy, thereby driving the anti-correlation between the slope and V-band attenuation, A_V. The scatter in this relation arises from differences in dust optical depth along and perpendicular to the line of sight, reflecting differences in the inclination and star-dust geometry. Additional contributions to the scatter come from variations in the grain size distribution and the fraction of obscured young stars.

Abstract Radiative feedback from massive stars plays a central role in the evolution of molecular clouds and the interstellar medium. This paper presents a multiwavelength analysis of the bright-rimmed cloud (BRC) 44, which is located at the periphery of the H ii region Sh2-145 and is excited by the massive stars in the region. We use a combination of archival and newly obtained infrared data, along with new optical observations, to provide a census of young stellar objects (YSOs) in the region and to estimate stellar parameters such as age, mass, etc. The spatial distribution of YSOs visible in the optical wavelength suggests that they are distributed in separate clumps compared to the embedded YSOs and are relatively older. Near-infrared spectroscopy of four YSOs in this region using the TANSPEC mounted on the 3.6 m Devasthal Optical Telescope confirms their youth. From spectral energy distribution fitting, most of the embedded YSO candidates are in their early stage of evolution, with the majority of them in their class II and some in their class I stage. The relative proper motions of the YSOs with respect to the ionizing source are indicative of the rocket effect in the BRC. The 12 CO, 13 CO, and C 18 O observations with the Purple Mountain Observatory are used to trace the distribution of molecular gas in the region. A comparison of the cold molecular gas distribution with simple analytical model calculations shows that the cloud is in the compression stage, and massive stars may be influencing the formation of young embedded stars in the BRC region due to radiative feedback.

Observations of Milky Way stars by multiplexed spectroscopic instruments and of gas in nearby galaxies using integral field units have made it possible to measure the abundances of multiple elements in both the interstellar medium and the stars that form out of it. These observations have revealed complex correlations between elemental abundances, but thus far there has been no analytic theoretical framework to interpret these data. In this paper we extend the simple stochastically-forced diffusion model of Krumholz & Ting (2018), which has proven successful at explaining the spatial abundance patterns of single elements, to multiple elements, clarifying why elements are correlated and what controls their degree of correlation, and making quantitative predictions for the degree of correlation in both gas and young stars. We show that our results are qualitatively consistent with observed patterns, and point out how application of this theory to measured correlations should enable determination of currently unknown parameters describing r-process nucleosynthesis.

In this paper, we explore the kinematic properties of a sample of 19 young ($<$1 Gyr) and intermediate-age (1--2.5 Gyr) massive star clusters within the Large Magellanic Cloud (LMC). We analysed the proper motions of the clusters, which have been measured based on multi-epoch Hubble Space Telescope (HST) observations. Additionally, from the HST data we inferred homogeneous and robust estimates for the distances, ages, and metallicities of the clusters. This collection of information, in combination with literature line-of-sight velocities, allowed us to investigate the full 3D dynamics of our sample of clusters within the frame of the LMC in a self-consistent way. While most young clusters orbit the LMC close to the stellar disc plane, NGC 1850 (∼100 Myr old) depicts a peculiar case. Depending on the exact distance from the disc, it follows either a highly inclined retrograde orbit or an eccentric orbit along the bar structure. The orbits of young clusters that formed north of the LMC centre show signs that might be connected to the resettling motion of the LMC bar structure. Based on the dynamic properties in combination with the positions of the clusters in the age-metallicity space, we find no clear-cut evidence of clusters in our sample that could have been stripped from the Small Magellanic Cloud (SMC) onto the LMC. Finally, we compared the kinematics of the intermediate-age clusters with a suite of simple numerical simulations of the Magellanic system to interpret the cluster motions. A possible interaction history of the LMC with the SMC, where the SMC had two past crossings of the LMC disc plane (about 300 and 900 Myr ago), in combination with the recent SMC pericentre passage, can qualitatively explain the observed kinematic structure of the clusters analysed in this work.

Abstract FU Ori outbursts are thought to play an important role in stellar assembly and the evolution of protoplanetary disks. However, the progenitor young stellar objects are largely uncharacterized. We obtained a low-resolution optical spectrum of HBC 722 before its FU Ori outburst as part of a survey of young stellar objects in the North America Nebula. The spectrum yields a spectral type of M3.3±0.4, which when combined with archival photometry allows us to measure the stellar and accretion properties of a young star prior to its FU Ori outburst. The pre-outburst accretion rate of 7 × 10 −9 M ⊙ yr −1 is high for a protoplanetary disk around an M3-M3.5 star, though about 15,000 times weaker than the accretion rate during the outburst. The pre-outburst variability, inferred from archival B -band photometry, is about a factor of 5 with a standard deviation of 0.16 dex and is consistent with variable accretion onto young low-mass stars. The stellar radius is larger than the radius of accreting young stars of similar spectral type by a factor of 2. The extinction to HBC 722 is ∼1.45 ± 0.3 mag, lower than the 2.5–3.7 mag extinction values measured during the outburst. The u -band photometry plays an especially important role in constraining the veiling at longer wavelengths and therefore also the extinction and photospheric luminosity.

Magnetic winds are a key mechanism for angular momentum removal in young stars. We aim to characterize the multi-component outflow of RU Lup, link discrete absorption components in resonance lines to forbidden-line emission, and quantify the mass loading, lever arms, and torque carried by the wind. The high resolution of the Echelle SPectrograph for Rocky Exoplanets and Stable Spectroscopic Observations (ESPRESSO) allowed us to perform a detailed study of the forbidden emission lines and the blueshifted absorption in the lines of the Na i and Ca ii doublets, which we resolved in three discrete absorption components at low, medium, and high velocities. We developed a method that disentangles vertical and toroidal velocities in the absorption components and infers the launching radius r_0, magnetic lever arm łambda, and dot M ̊m wind We identify a low-velocity broad component in the O i 5577 line, consistent with a rotating magnetohydrodynamic disk wind launched near the disk truncation radius. We show that the discrete absorption components trace spatially and physically distinct regions of the outflow. The high-velocity absorption component is connected to the high-velocity component of the forbidden lines and it is formed in a knot in the large-scale, low-density jet. The medium- and low-velocity components are launched from the inner disk (r_0 łesssim 6.76,R_⋆) with low lever arms indicative of warm, highly mass-loaded streamlines. The two components differ mainly in their vertical velocity. The low-velocity absorption is consistent with an outer absorbing shell, whereas the medium-velocity absorption forms near the disk truncation radius. Its higher vertical velocity is compatible with either a slightly larger lever arm or additional heating at the base of the flow. For plausible ionization levels in the inner disk, this outflow component removes a substantial fraction of the accretion spin-up torque. RU Lup hosts a stratified, rotating, warm disk wind launched across a narrow annulus near the disk truncation radius, which is sufficiently mass-loaded to extract a large fraction of the stellar spin-up torque. These observations disfavor an X-wind scenario.

Abstract Our knowledge of the chemical composition of the gas in the inner disc of intermediate-mass young stars is limited, due to the lack of suitable instrumentation. The launch of JWST has provided a significant improvement in our ability to probe gas in these inner discs. We analyse the gas composition and emitting conditions of the disc around HD 35929, a young intermediate-mass Herbig star, using MIRI/MRS data. Our goal is to constrain the chemistry and kinematics of the gas phase molecules detected in the inner disc. We use iSLAT to examine the observed molecular lines and DuCKLiNG to detect, fit, and analyse the molecular emission. We find gas phase H2O, CO, CO2, and OH in the disc, as well as HI recombination lines. Surprisingly, we also detect gas phase SiO in the fundamental v=1-0 vibrational band. We derive column densities and temperature ranges of the detected species, arising from the inner $\sim 0.2\, \rm au$, hinting towards a compact and very warm disc. The molecular column densities are much higher than found in lower mass T Tauri discs. In general, the molecular composition is consistent with an O-rich gas from which silicate-rich solids condense and the strong gas phase molecular line emission suggests a low dust opacity. The unexpected detection of gas phase SiO at the source velocity points to an incomplete condensation of rock forming elements in the disc, suggesting chemical disequilibrium and/or an underestimate of the gas kinetic temperature.

Abstract Stellar astrophysics relies on diverse observational modalities—primarily photometric light curves and spectroscopic data—from which fundamental stellar properties are inferred. While machine learning (ML) has advanced analysis within individual modalities, the complementary information encoded across modalities remains largely underexploited. We present the dual embedding for stellar astronomy (DESA) model, a novel multimodal foundation model that integrates light curves and spectra to learn a unified, physically meaningful latent space for stars. DESA first trains separate modality-specific encoders using a hybrid supervised/self-supervised scheme, and then aligns them through DualFormer, a transformer-based cross-modal integration module tailored for astrophysical data. DualFormer combines cross- and self-attention, a novel dual-projection alignment loss, and a projection-space eigendecomposition that yields physically structured embeddings. We demonstrate that DESA significantly outperforms leading unimodal and self-supervised baselines across a range of tasks. In zero- and few-shot settings, DESA’s learned representations recover stellar color–magnitude and Hertzsprung–Russell diagrams with high fidelity ( R 2 = 0.92 for photometric regressions). In full fine-tuning, DESA achieves state-of-the-art accuracy for binary star detection (AUC = 0.99, AP = 1.00) and stellar age prediction (RMSE = 0.94 Gyr). As a compelling case, DESA naturally separates synchronized binaries from young stars—two populations with nearly identical light curves—purely from their embedded positions in UMAP space, without requiring external kinematic or luminosity information. DESA thus offers a powerful new framework for multimodal, data-driven stellar population analysis, enabling both accurate prediction and novel discovery.

Abstract In the early stages of their lives, young stars harbor diverse processes exhibiting strong signatures, making them ideal targets for investigating stellar accretion, dynamo actions, and exoplanet formation. However, the current samples of active stars are highly limited in both number and parameter space, suffering from significant statistical uncertainties. By simultaneously measuring the activity proxy, the Balmer H α 6563 Å line emission, and the youth indicator, the lithium 6708 Å line, with the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) spectra, we have identified more than 1000 candidates of active young stars (aYSs) with T eff ≤ 7000 K and log g ≥ 3.5 in the Kepler prime field. Then we matched 660 stars with the Kepler light curves, and determined their periods and variable types. Based on these light curves, we are left with 78 dubious and 52 stable stars, and we have identified 22 eclipsing binaries, 3 heartbeat binaries, 1 transiting exoplanetary system candidate, 13 long-period variables, 61 pulsating stars, 18 irregular variables, 57 T Tauri stars, 173 flaring stars, and 435 rotational variables (the numbers are not exclusive). Supplementing the cluster-based samples, this sample has a broader parameter space coverage. This work will be beneficial for identifying aYSs in much larger datasets, such as the one produced by crossmatching between the LAMOST and Transiting Exoplanet Survey Satellite surveys, and the future survey conducted by the Chinese Space Station Survey Telescope.

ABSTRACT The angular momentum evolution of stars is crucial for understanding the formation and evolution of stars and star clusters. Using high-resolution magnetohydrodynamical simulations of star formation in clouds with different physical properties, we study the initial distribution of stellar rotation periods in young clusters. We compare these results with observations of young Galactic clusters. Simulations qualitatively reproduce the observed trend of increasing rotation period with stellar mass. Additionally, simulations with lower virial parameter (ratio of turbulence to gravity) or solenoidal turbulence driving produce period-mass distributions that more closely match the observed ones. These simulations also recover the break in the mass-period relation. However, the break appears at higher masses than in observations and is absent in the youngest simulated clusters. This suggests that the emergence of the break is an important diagnostic of angular momentum evolution during the earliest stages of cluster formation. The simulations yield stars that rotate about an order of magnitude faster than those observed. This discrepancy mainly reflects the earlier evolutionary stage of the simulations, while unresolved physical interactions between stars and discs might also contribute. This conclusion is supported by simulations showing a significant period increase within $0.1{\!-\!}1\, \mathrm{Myr}$. We quantify the required angular-momentum loss by rescaling simulated rotation periods to match observations, finding that $80{\!-\!}95~{{\ \rm per\, cent}}$ of the initial angular momentum must be removed within the first $\mathrm{Myr}$. Our results highlight that understanding the earliest stages of star cluster formation is fundamental to addressing the angular momentum problem.

We present simulations of a massive young star cluster using the codes Nbody6++GPU and MOCCA . The cluster is initially more compact than previously published models. It contains one million stars and has a total mass of $5.86 M _⊙$ and a half-mass radius of $0.1 pc We analyzed the formation and growth of a very massive star (VMS) through successive stellar collisions and investigated the subsequent formation of an intermediate-mass black hole (IMBH) in the core of a dense star cluster. We used direct N -body and Monte Carlo simulations that incorporated updated stellar evolution prescriptions for single and binary stellar evolution (SSE and BSE) tailored to massive stars and VMSs. These include revised treatments of stellar radii, rejuvenation, and mass loss during collisions. While the prescriptions represent reasonable extrapolations into the VMS regime, the internal structure and thermal state of VMSs that formed through stellar collisions remain uncertain, and future work may require further refinement. Runaway stellar collisions in the cluster core produce a VMS that exceeds $5 M _⊙$ within 5 Myr that subsequently collapses into an IMBH. We stress that further work on stellar astrophysics is needed, particularly in the context of VMS formation. The VMS formation currently represents strong uncertainties. Our model suggests that dense stellar environments may enable the formation of VMSs and massive black hole seeds through runaway stellar collisions. These results provide a potential pathway for early black hole growth in star clusters and offer a theoretical context for interpreting recent observations with the James Webb Space Telescope of young compact clusters at high redshift.

GRAVITY+ improves by orders of magnitude the sensitivity, sky-coverage, and contrast of the Very Large Telescope Interferometer (VLTI). A central part of this project is the development of Gravity Plus Adaptive Optics (GPAO), a dedicated high-order and laser-guide star adaptive optics (AO) system for VLTI. GPAO consists of four state-of-the-art AO systems that equip all 8m class Unit Telescopes (UTs) for the wavefront correction of the VLTI instruments. It offers both visible and infrared natural guide star (NGS) and laser guide star (LGS) operations. The paper presents the design, operations, and performances of GPAO. We illustrate the improvement brought by GPAO with interferometric observations obtained during the commissioning of the NGS mode at the end of 2024. These science results include the first optical interferometry observations of a redshift z∼4 quasar, the spectroscopy of a cool brown-dwarf with magnitude K∼ 21.0, the first observations of a Class I young star with GRAVITY, and the first sub-micro arcsecond differential astrometry in the optical. Together with the entire GRAVITY+ project, the implementation of GPAO is a true paradigm shift for observing the optical Universe at very high angular resolution.