The standard interpretation of observed Stokes profiles to infer the physical conditions of the solar atmosphere is inherently an ill-defined problem due to observational noise and mathematical degeneracies. Traditional pixel-by-pixel (1D) inversion codes provide point estimates with unreliable uncertainties, at the expense of significant computational time. Recent machine-learning-based Bayesian frameworks are restricted to 1D spatial configurations, ignoring crucial spatial correlations between neighboring pixels. We aim to develop a novel multidimensional inversion framework capable of performing fast and scalable Bayesian inference across an entire 2D field-of-view (FoV). This approach seeks to provide accurate height-dependent atmospheric parameters with reliable posterior distributions while exploiting spatial correlations. We introduce a new generative modeling strategy based on conditional flow matching. The model utilizes multi-scale spatial features extracted from observed Stokes profiles in the Fe I line pair at 630 nm, which then conditions a flow matching generative model to sample from the complex posterior distribution of the atmospheric parameters. The framework is trained using realistic 3D quiet Sun magnetohydrodynamic simulations. Validation on independent synthetic datasets demonstrates that the model accurately captures the true 3D stratification of all thermodynamic and magnetic parameters. Because the code additionally provides a geometrical height scale, it allows for the computation of 3D electric current density maps, Lorentz forces, and Ohmic and ambipolar dissipation maps in the solar photosphere. Application to real Hinode/SP quiet Sun observations yields highly localized electric currents at magnetic boundaries. We also leverage the 3D geometrical information to trace the emergence of small-scale emerging magnetic loops across the solar atmosphere.

The $\Lambda$CDM cosmological model has been successful in explaining many astronomical observations. However, recent observations increasingly point to deviations from the standard $\Lambda$CDM framework. Among these, one of the most significant discrepancies is the \textit{Hubble tension}, which refers to the difference in values obtained for the Hubble constant $H_0$ from high-redshift measurement and local observation. To address this issue, numerous cosmological models and methodological approaches have been proposed. This review offers a concise overview of recent progress in resolving the Hubble tension. The combination of Dark Energy Spectroscopic Instrument (DESI) Baryon Acoustic Oscillations (BAO) and uncalibrated Type Ia supernovae data yields a value for $H_0$ that is significantly higher than the $\Lambda$CDM predication based on early-universe probes, even without incorporating local distance ladder constraints. This result indicates that the origin of the Hubble tension lies in new physics at low redshifts. Our findings suggest that although many unresolved systematics persist in current observations, they are insufficient to account for the magnitude of the current Hubble tension. This implies the likely existence of new physical mechanisms that have yet to be discovered.

Fast radio bursts (FRBs) are millisecond-duration radio flashes of unknown origin, observationally classified into repeating and apparently nonrepeating (one-off) populations. In this work, we use a statistical population approach to investigate the redshift evolution of one-off FRBs. We compare a pure star formation history (SFH) tracing model, phenomenological delayed models, physically motivated delayed models that correspond to binary neutron star related and neutron star age-window channels, and mixture models which is obtained when two physically motivated models are normalized separately and then combined as a weighted mixture. The samples in CHIME/FRB Catalog do not support an intrinsic event rate density that directly follows the SFH. The preferred model is mixture model corresponds to an effective mean delay time of $\bar{\tau}=1.426^{+0.032}_{-0.035}~\mathrm{Gyr}$. These results suggest that the current data may naturally explained by delayed, possibly multi-channel progenitor evolution for the one-off FRBs.

The gravitational wave background (GWB) produced by extreme-mass-ratio inspirals (EMRIs) serves as a powerful tool for probing the astrophysical and dynamical processes in galactic centers. EMRI systems are a primary target for the space-based detector LISA due to their long-lived signals and high signal-to-noise ratios. This study explores the statistical properties of the GWB from EMRI, focusing on the calculation methods for the GWB, the astrophysical distribution of EMRI sources, and the influence of key parameters, including the spin of supermassive black holes (SMBHs) and the masses of compact objects (COs). By analyzing these factors, we determine the distribution range of the characteristic strain of the GWB from EMRIs. We find that the final eccentricity distributions appear to have negligible effect on the intensity of the GWB due to rapid circularization before they become detectable and the spin of the SMBH enhances the GW characteristic strain by approximately 1$\%$ compared to cases without spin effects. The masses of COs can also significantly affect the characteristic strain of the GWB from EMRIs, with Black Hole (BH) as CO producing a GW signal intensity that is approximately one order of magnitude higher compared to cases where Neutron Star (NS) or White Dwarf (WD) are the COs.

The masses of galaxy clusters carry cosmological and astrophysical information. We develop a simulation-based inference pipeline to infer cluster masses from full projected phase-space information of member and interloper galaxies. Our method combines a permutation-invariant Deep Sets architecture with neural posterior estimation using normalizing flows, enabling the recovery of expressive posterior distributions. We train the model to predict residual corrections to the classical $M$--$\sigma$ relation, thus explicitly isolating information beyond velocity dispersion. Using the Uchuu-UniverseMachine simulation, we evaluate the method under both idealized (interloper-free) and realistic (cylindrical) observational setups. In the idealized case, our model reduces the scatter in mass estimates to as low as $\sim 0.1$ dex, representing a twofold improvement over the traditional $M$--$\sigma$ relation. In the cylindrical setup, we achieve comparable performance at the high-mass end ($> 10^{14.5}\,M_\odot/h$), demonstrating robustness against interloper contamination. We demonstrate that set-based simulation-driven inference provides a powerful and flexible framework for galaxy cluster mass estimation, enabling improved accuracy and reliable uncertainty characterization for upcoming large-scale surveys. Our model saturates the kinematic information content and thus suggests a baseline for future studies.

A $10^6\,M_\odot$ object has recently been inferred from gravitational imaging of the strong-lensing system JVAS B1938+666, exhibiting an unusually dense inner region embedded within an extended envelope, far exceeding expectations for cold dark matter (CDM) halos. Using gravothermal fluid simulations, we show that such a structure arises naturally in self-interacting dark matter (SIDM) halos evolving into a deep core-collapse phase, where a secondary dense central core forms within an extended profile. The resulting density structure closely matches the inferred properties of the lensing object. We also demonstrate that a similar profile could be reproduced in CDM in the presence of an intermediate-mass black hole, but this requires an early-forming progenitor that subsequently loses $5$ orders of magnitude in mass through tidal stripping by the lens galaxy. Whether such a scenario can be realized in realistic cosmological environments remains an open question.

Many details of the gas accretion phase during giant planet formation remain untested. We present new 0.2$\unicode{x2013}$0.7 $\mu$m UV-through-red optical imaging of the young, wide-orbit planetary-mass companion SR 12 c from the Wide Field Camera 3 (WFC3) instrument on board the Hubble Space Telescope. SR 12 c exhibits strong accretion-related continuum excess blueward of $\sim$5000 $\unicode{x212B}$ and clear signs of the Balmer jump at 3646 $\unicode{x212B}$. We derive a total accretion luminosity of 1.65 $\pm$ $0.19 \times 10^{-5} L_{\odot}$ and a mass accretion rate of 8 $\pm$ $2\times 10^{-12}$ M$_{\odot}$ yr$^{-1}$. Based on its mass and age, SR 12 c will not grow by an appreciable amount at its current accretion rate; it is at the end stages of assembly. No accretion variability is evident between the two epochs of the WFC3 observations spanning a month-long baseline, but the H$\alpha$ emission line strength decreases by 90% compared to the reported flux from five years earlier. Combined with previous observations of SR 12 c, we assemble one of the most complete spectral energy distributions of a young giant planet to date, spanning the UV through sub-mm wavelengths (0.2$\unicode{x2013}$880 $\mu$m). This adds SR 12 c to the small yet growing sample of planets with detailed accretion and disk constraints, which together are beginning to establish the diversity of timescales and physical processes governing the formation of giant planets.

The JWST discoveries of supermassive black holes (BHs) at $z \gtrsim 5$ may provide key insights into their seeding origins. Using new $[18{-}72~\rm Mpc]^3$ BRAHMA cosmological simulations, we investigate how variations in heavy-seed prescriptions, coupled with a subgrid dynamical friction model, shape BH populations at $z \sim 5$ and $z \sim 0$. We consider two "lenient'' seed models, in which all halos containing sufficient dense & metal-poor gas form $\sim10^4$ and $\sim10^5~M_{\odot}$ seeds, and a "strict'' seed model, in which $\sim10^5 M_{\odot}$ seeds form only under additional constraints motivated by direct collapse black hole formation. By $z \sim 5$, all models produce $M_*-M_{\rm BH}$ relations broadly consistent with the observed local Universe for $M_*\gtrsim10^9~M_{\odot}$ galaxies, but only the lenient scenarios generate systems near the upper envelope of the observed local scatter. In galaxies hosting $M_{\rm BH} \sim 10^8$-$10^9~M_{\odot}$ BHs, lenient production of $\sim10^5~M_{\odot}$ seeds also produces multiple overmassive systems with $M_{\rm BH}/M_* \gtrsim 0.01$. Although their growth is dominated by seeding and mergers, these systems reach luminosities of $\sim10^{43}$-$10^{45}\mathrm{erg s^{-1}}$, comparable to those inferred for JWST-detected BHs. As a key observational signature, the lenient seed models yield merger rates of $\gtrsim100\mathrm{yr^{-1}}$ and near-unity local BH occupation fractions even in galaxies with $M_* \lesssim 10^7~M_{\odot}$. In contrast, the strict seed model produces merger rates of only $\sim1\mathrm{yr^{-1}}$ and local occupation fractions of $\lesssim10\%$ for galaxies with $M_* \lesssim 10^8~M_{\odot}$. Future gravitational-wave event rates and measurements of local BH occupation fractions will therefore provide strong constraints on the dominant pathways responsible for high-redshift BH assembly.

Synchrotron emission in high-energy transients is produced by relativistic electrons accelerated by shocks. As high-energy transients are often unresolved even on angular scales probed by very long baseline interferometry, it is difficult to obtain a full picture of the ejecta and circumstellar medium (CSM) properties that are probed by the radio synchrotron emission. Radio spectra of high-energy transients frequently show optically thick slopes shallower than the standard $F_\nu \propto \nu^{5/2}$ expected from synchrotron self-absorption (SSA) models, or broader spectra near the self-absorption frequency. Such deviations are often interpreted phenomenologically, without providing clear insights into the structure of the emitting region. We show how information on the homogeneity and symmetry of the emitting region can be directly inferred from SSA spectra, even when the source is unresolved. We discuss the circumstances under which inhomogeneities in the emitting region can change the spectrum below the self-absorption frequency, causing it to follow a different slope. We examine which parameters can be constrained from observations and which remain degenerate. We apply this method to the stripped-envelope supernova (SN) 2016coi and to the fast blue optical transient (FBOT) AT2018cow, showing that SSA spectra constrain the degree of inhomogeneity in these systems, providing strong evidence for inhomogeneities in the emitting region in the SN 2016coi, and asymmetry in the case of AT2018cow, and we infer the characteristics of the emitting region. When well sampled spectra are available, our method can be applied as a general, model-independent, inference method. This approach can be used to constrain inhomogeneities in a variety of unresolved high-energy astrophysical transients, including SNe, FBOTs, tidal disruption events and gamma-ray bursts.

The Habitable Worlds Observatory (HWO) aims to enable the detection and characterization of Earth-like planets around Sun-like stars to search for possible signs of life elsewhere in our universe. This requires an incredibly sensitive coronagraph instrument that suppresses the light from the star by a factor of 10 billion, which must contend with error terms that have not previously limited high-contrast instrumentation at lower levels of starlight suppression. Polarization aberrations are one such source of error that is particularly problematic for coronagraphy on a large space telescope. Optical rays in large, compact astronomical observatories can have large changes in angle of incidence over the beam, which induce polarization aberrations that decrease sensitivities to faint signals at small angular separations. Limiting variation in angles of incidence along the optical path could lead to longer, less stable observatories. This could negatively impact the total number of exo-Earths HWO would be able to detect. This study links open-source physical optics modeling tools to an exoplanet yield optimizer to understand how polarization aberrations influence science return for HWO. We also explore how polarization aberrations scale with change in angle of incidence, which could drive the primary-secondary mirror distance and overall observatory stability. In the visible, we find that decreasing the EAC-1 barrel from 16m to 12m results in $\approx 10^{-10}$ contrast at the IWA where we expect exo-Earths to be. In the UV we appear to be less sensitive to polarization because exo-Earths are farther from the IWA. We also find a limited range over which the design reference mission of EAC-1 can be optimized to compensate for polarization aberrations using altruistic yield optimization. We then report on mitigation strategies to minimize the presence of polarization aberrations in HWO.

We present a data-driven investigation of Galactic diffuse emission. Using multi-frequency Planck radio/microwave maps (30-857 GHz) and Fermi-LAT gamma-ray data (50 MeV-814 GeV), we construct a nonlinear mapping between radio emission and gamma-ray intensity through supervised machine learning. Our models achieve high predictive accuracy (R^2 > 0.90 in the 0.1-10 GeV range), demonstrating that multi-frequency radio observations encode sufficient information to reconstruct both spatial morphology and spectral properties of diffuse gamma-ray emission. By analyzing model performance across different frequency bands and spatial regions, we identify high-frequency radio bands as the dominant predictor, providing direct empirical support for the hadronic origin of Galactic 0.1-10 GeV gamma rays, while low-frequency radio bands for the leptonic origin above 10 GeV. Residual maps reveal coherent large-scale structures, including Loop I and III, highlighting regions where standard interstellar emission models are incomplete or biased. Compared with the GALPROP model, our machine learning approach yields a higher R^2=0.95 and lower mean absolute relative error (14.7%) in the inner Galactic disk and the Galactic center region. Our results illustrate that machine learning serves as a physically interpretable tool for multi-messenger astrophysics, providing a data-driven baseline for separating non-standard emission components and deriving new constraints on cosmic-ray propagation and interstellar medium structure.

We identify a geometric obstruction to resolving the Hubble tension by combining early-time sound-horizon reduction with late-time smooth dark energy. Within $\Lambda$CDM, the BAO--SN matter-density gap $\Delta\Omega_m = 0.037$ is exactly invariant under the sound-horizon rescaling $\alpha \equiv r_s^{\rm mod}/r_s^{\Lambda{\rm CDM}}$, and late-time $w(z)$ deformations cannot eliminate this gap either: reconciling the two datasets requires \emph{opposite} deformations -- phantom ($w < -1$) for BAO, quintessence ($w > -1$) for SN at $z < 0.5$ -- an anti-alignment quantified by $\cos\theta = -0.97$ in $w(z)$ space. A full MCMC analysis of DESI DR2 BAO, Planck plik\_lite, and Pantheon+ bears this out: the optimal $\alpha^* = 0.992$ ($0.8\%$ $r_s$ reduction) brings the joint fit to $H_0 = 70.3 \pm 0.3\;\mathrm{km\,s^{-1}\,Mpc^{-1}}$, still $3.2\sigma$ below SH0ES, with the inter-dataset tension reduced but not removed. The obstruction reflects not a shortage of model freedom but an irreducible disagreement between probes. The deformation space $\{\alpha, \beta_{\rm damp}, w(z)\}$ already spans $93\%$ of the $\Omega_m$ response direction; nonetheless BAO and SN constrain $\Omega_m$ through independent channels and disagree, while the residual $H_0$ deficit, anchored by the local distance ladder, resides in the absolute distance scale that $w(z)$ reshapes but cannot rescale.

Strong lensing of massive black hole binaries can produce multiple gravitational-wave images with different magnifications and arrival times. LISA signals remain in band for months to years, allowing multiple lensed images to overlap during the inspiral stage and generate beat patterns. A singular isothermal sphere lens model is adopted to describe the lensing configuration, and two-image beat waveforms are constructed from massive black hole binary signals. To isolate the beat pattern itself, waveform mismatch is evaluated only during the overlapping inspiral stage before the coalescence of the first image, excluding contributions from the delayed merger peak of the second image. Using the HS-nod-SN (B+20) strong-lensing population, the occurrence rate of identifiable beat events is estimated, and Bayesian parameter estimation is performed with a beat template. Beat patterns are most readily identified when the lensing time delay is short and the delayed image has a relatively large magnification. Among 196 detectable two-image lensed events, 92 satisfy the temporal-overlap condition and 14 satisfy the beat-identification criterion, corresponding to an identifiable beat fraction of about 7\%. Posterior inference shows that the beat template can recover the lensing time delay and magnification parameters for a representative beat event. These results indicate that lensed beat patterns constitute a distinguishable subset of strongly lensed LISA events and provide a unique observational signature of strong lensing in the LISA band.

We present an analysis of XMM-Newton and Chandra observations of 85 nearby main-sequence FGK stars with age estimates ranging from 0.2-12 Gyr. We measure quiescent 0.3-10 keV luminosities, variability metrics, and multi-temperature thermal plasma spectral parameters. Quiescent spectra are typically described by three characteristic plasma components ($kT\approx0.1$, 0.4, 0.8 keV); the fraction of flux from $T\ge7$ MK rises with X-ray surface flux, reaching $\sim$50% for $F_X\gtrsim10^6$ erg cm$^{-2}$ s$^{-1}$. We derive relations between emission measure-weighted coronal temperature and both $L_X$ and $F_X$, enabling temperature-informed count-rate conversions for faint sources. We quantify how bandpass conversions (ROSAT 0.1-2.4 keV vs. XMM-Newton 0.3-10 keV) depend on temperature, and show that inferred ROSAT-band $L_X$ broadly follows the canonical $t^{-1.5}$ decay, while the harder band exhibits increased scatter at $>$4 Gyr. Several stars show excess activity suggestive of age errors, inclination effects, or unresolved companions. Some of these "outlier" stars are potential direct imaging targets for the Habitable Worlds Observatory, and detailed characterization of these stars is needed to inform their likely influence on the atmospheric evolution of orbiting planets.

High-mass X-ray binaries (HMXBs) provide a natural laboratory to study radiatively driven stellar winds under strong external X-ray irradiation. As the compact object moves along its orbit, the wind density and ionization structure vary with orbital phase, leaving characteristic signatures in X-ray emission and absorption features. The amplitude and morphology of this variability depend on the system geometry, including the orbital inclination (via line-of-sight and occultation effects) and the orbital eccentricity (via phase-dependent changes in the orbital separation). We present a computational framework that connects phase-resolved spectroscopic variability to the three-dimensional structure of irradiated winds, and that enables inference of physically meaningful wind--irradiation parameters within a Bayesian setting. We combine photoionization calculations with a parametric wind description and orbital geometry to construct three-dimensional maps of density and ionization. From these maps we compute orbital-phase-dependent diagnostics, accounting for geometric occultation and wind inhomogeneity through a clumping prescription. We then use Bayesian inference to compare model predictions with phase-resolved observables and to quantify parameter constraints and degeneracies. The framework reproduces the main orbital-phase-dependent trends expected for irradiated winds and yields robust constraints on the system ionization balance. While combinations of wind and luminosity parameters are well constrained, individual parameters can remain partially degenerate depending on the orbital configuration and data quality. This modular and computationally efficient approach provides a route to interpret HMXB phase variability in physical terms, and offers a foundation for future extensions toward forward spectral modeling and population-level applications.

We investigate how low-ionization gas flows in typical star-forming galaxies at $z\sim3$ depend on galaxy intrinsic properties and viewing angle. For this analysis we use JWST/NIRSpec observations of rest-frame near-UV Fe II and Mg II absorption, and rest-frame optical Na D absorption. This study combines galaxies from the LyC22, EXCELS, and AURORA surveys and contains 176, 197, and 315 galaxies, respectively, with Fe II, Mg II, and Na D coverage. Based on both individual and composite spectra, we find no statistically significant correlations between outflow velocity and galaxy properties. However, galaxies with detected outflows tend towards higher stellar masses, SFR, and $\Sigma_{\rm SFR}$ than those without outflows, suggesting that the two samples are not drawn from the same parent population. Finally, we additionally find that Mg II emission is preferentially detected in galaxies with lower stellar mass and $A_V$, and higher sSFR, consistent with conditions that favor the escape of resonantly scattered line and ionizing continuum radiation. We present the first evidence in $z\sim3$ star-forming galaxies that properties of the absorption lines depend on galaxy inclination, with more face-on systems showing stronger absorption and higher outflow velocities, while inflowing gas is more frequently detected in more highly inclined galaxies. These trends are consistent with observations at $z\lesssim1$ and predictions from cosmological simulations in which galactic winds are launched perpendicular to the galactic disks, while accretion occurs primarily along the disk plane.

Using a sample of 151 galaxies with dynamically measured black hole (BH) masses ($M_{\rm BH}$), we investigate the scaling relations between $M_{\rm BH}$ and the stellar velocity dispersion, $\sigma$, and, for a subsample of 30 core-Sérsic galaxies, between $M_{\rm BH}$ and the size of the partially depleted core, $R_{\rm b}$. Core-Sérsic galaxies, identified using high-resolution $Hubble ~ Space ~ Telescope$ imaging and spanning both the normal-core $(R_{\rm b}<0.5$ kpc) and large-core ($R_{\rm b}>0.5$ kpc) regimes, define an updated $M_{\rm BH}$$-$$R_{\rm b}$ relation of the form $M_{\rm BH} \propto R_{\rm b}^{1.16 \pm 0.10}$, with an rms scatter of $\Delta_{\rm rms} \simeq 0.28$ dex in $\log M_{\rm BH}$. We find that Sérsic and normal-core galaxies together follow a common log-linear $M_{\rm BH}$-$\sigma$ relation with a slope of $4.95 \pm 0.29$ and a scatter $\Delta_{\rm rms} \simeq 0.46$ dex. Deviations from this relation arise at the highest BH masses, where large-core galaxies, including six with direct $M_{\rm BH}$ measurements, drive a significant upturn. We find that these galaxies typically host ultramassive black holes whose masses scale more strongly with $R_{\rm b}$ than $\sigma$, and lie $\sim (1-4$) $~\times$ the intrinsic scatter (0.39 dex) above the relation defined by Sérsic and normal-core galaxies. The $M_{\rm BH}$$-$$R_{\rm b}$ relation shows $\sim 30$-$47\%$ less scatter in $\log M_{\rm BH}$ than the corresponding $M_{\rm BH}$$-$$\sigma$ relation for the same sample. We interpret the high-mass upturn in the $M_{\rm BH}$$-$$\sigma$ diagram as a consequence of successive major, dry mergers, a scenario that naturally explains the observed flattening of the $\sigma-L_V$ relation at $M_V < -23.5$ mag.

This study presents a comprehensive X-ray spectroscopic analysis of 642 quasars, obtained by cross-matching the XMM-Newton Serendipitous Source Catalog (DR11) with the Sloan Digital Sky Survey (DR16) quasar catalog. After stringent quality filtering and automated spectral reduction, we derived reliable photon indices ($\Gamma$) and intrinsic 2--10 keV X-ray luminosities. Using multiwavelength data, sources were classified into 561 radio-quiet (RQ) and 81 radio-loud (RL) quasars. We estimate the bolometric luminosity and Eddington ratio ($\lambda_{\mathrm{Edd}}$) from absorption-corrected X-ray measurements and virial black hole masses. Our primary objective is to establish and characterize the fundamental relationship between the photon index and Eddington ratio for RQ population. We find that RQ quasars exhibit systematically higher Eddington ratios, peaking at $\log \lambda_{\mathrm{Edd}} \approx -1.2$, and softer spectra with $\Gamma \approx 2.0$. A statistically significant positive correlation between $\Gamma$ and $\lambda_{\mathrm{Edd}}$ is detected in RQ quasars, supporting disk--corona coupling models. To validate our results within the broader context of AGN evolution, we further examine the dependence of $\lambda_{\mathrm{Edd}}$ on redshift ($z$) and black hole mass ($M_{\mathrm{BH}}$). For RQ quasars, $\lambda_{\mathrm{Edd}}$ increases with redshift and decreases with $M_{\mathrm{BH}}$, in strong agreement with recent results \cite{aggarwal2024evidence}, highlighting the universal nature of these accretion trends. By correlating spectral slope with accretion rate, this work provides new insights into the interplay between accretion physics, jet activity, and the cosmic evolution of quasars.

Solar wind stream interaction regions (SIRs) drive recurrent geomagnetic storms, yet most existing catalogs rely on expert inspection and simple thresholds that are subjective and can miss events with complex morphologies. We present SIREN (SIR Encoder Network), a lightweight Transformer based model for per timestep SIR detection from in situ solar wind observations. The model ingests sequences of 11 solar wind parameters, spanning magnetic field, velocity, and thermodynamic properties. With approximately 100,000 trainable parameters in a two layer encoder architecture, SIREN is trained using weighted binary cross entropy loss and a cosine annealing learning rate. Platt scaling is applied to produce well-calibrated detection probabilities. On a held-out test set of 102 events, the calibrated model achieves a ROC-AUC of 0.93, F1 score of 0.78, and true skill statistic of 0.67. Analysis of the self-attention weights confirms that the model concentrates on the SIR, grounding its decisions in the physically relevant portion of each sequence. Integrated Gradients attribution reveals a quantifiable feature hierarchy: proton density (24.3%) and magnetic field magnitude (21.6%) dominate, followed by temperature (13.9%) and bulk speed (12.1%). Notably, the transverse velocity component Vy and east-west flow angle together contribute 13-17%, identifying flow deflection as a consistent but previously under-quantified SIR signature. By producing continuous probabilities rather than binary labels, SIREN enables flexible threshold tuning for operational use and provides a template for compact, interpretable deep-learning systems in space weather.

This paper describes lessons learned over a long career building astronomical instruments. Although these lessons are based on one person's experiences, they were learned while working on major and minor instrument projects with many scientific and engineering colleagues. When I interact with colleagues entering the field, I am reminded that the "obvious" approaches were in fact learned over many years by observing others develop instruments, by receiving good advice, and by making mistakes. I hope to help others avoid making all of the same mistakes.

The OSIRIS-REx mission acquired spatially resolved (2-10 m spot sizes) visible-near infrared (VNIR) and thermal infrared (TIR) spectra across four candidate sampling sites on asteroid (101955) Bennu: Nightingale, Osprey, Sandpiper, and Kingfisher. To quantify heterogeneity across a small body (about 500 m radius) like Bennu, we explore remotely observed spectral data to draw conclusions about the mineralogical composition and key physical processes that drive surface variability. We derive diagnostic band parameters from the OSIRIS-REx Visible and Infrared Spectrometer and the OSIRIS-REx Thermal Emission Spectrometer datasets to quantify compositional and physical variability across sites and assess their mineralogical context. The VNIR spectra exhibit similar overall reflectance shapes but systematic differences in spectral slopes and the 2.74 micron OH absorption. TIR emissivity spectra reveal modest but statistically significant shifts in the Christiansen Feature, silicate stretching, and bending band positions, indicating differences in silicate composition, hydration state, and Mg/Fe relative abundance. Principal component analysis separates each site into distinct clusters in multivariate band-parameter space, whereas K-means clustering identifies intra-site spectral sub-populations. Welch's Analysis of Variance and Hotelling's tests confirm that band-parameter variations between sites are significant. These results reveal that Bennu's surface preserves measurable spectral heterogeneity at 2-10 m scales, with site-to-site variations in hydration indicators and silicate band positions. The spectral properties of Nightingale encompass the full range observed across all four sites, establishing a remote sensing baseline for contextualizing laboratory analyses of the returned sample within Bennu's broader composition diversity and alteration history.

Carbon-enhanced metal-poor (CEMP; with $\rm{[Fe/H]} \le -2.0$ and $\rm{[C/Fe]} \ge 0.7$) stars preserve information about early chemical enrichment, low-mass star formation, and the hierarchical assembly of galaxies. In this study, we have compiled an extensive literature sample of 1032 stellar carbon abundances spanning the metal-poor Milky Way halo (437 stars), 21 ultra-faint dwarf galaxies (UFDs; 102 stars), seven classical dwarf spheroidal galaxies (254 stars), three accreted dwarf galaxies (90 stars), the Small Accreted Stellar Systems (SASS; 77 stars), and eleven stellar streams (72 stars). We establish the fractions of CEMP stars for each of these systems and categories. Generally, the low-mass UFDs possess the high fractions at low metallicities, whereas the more massive classical dwarf galaxies have relatively few CEMP stars. This behavior reveals a new low-metallicity Magnitude ($M_{\rm V}$)--CEMP Fraction relation across the dwarf satellite galaxy population. The high CEMP fractions in surviving UFDs suggest their enrichment was dominated by faint supernovae, as higher energy input would likely have quenched star production. The low CEMP fractions in classical dwarfs imply predominantly in situ formation rather than assembly from smaller systems. Using $\rm{[C/H]}$ abundances, we also probe early low-mass star formation. Eight stars lie within or near the theoretical ''forbidden zone'', indicating that dust-induced cooling, alongside fine-structure line cooling, contributed to early star formation. These rare dust-cooled stars may have formed in UFD-like systems that did not survive. Overall, the metal-poor Milky Way halo appears to have assembled from many different dwarf galaxies, with CEMP halo stars being contributed by early UFD-like systems and non-CEMP halo stars by intermediate-sized halos that later formed classical dwarfs.

We describe spectroscopic tomographic weak lensing measurements (spectrotomography) for two rich clusters of galaxies, A1767 and A2065, based on extensive spectroscopy and Subaru/Hyper Suprime-Cam (HSC) imaging. These detections represent the first use of spectrotomography based on archival Subaru/HSC imaging. The measurements depend only on galaxies with spectroscopic redshifts reported here. The approach cleanly separates cluster members from the background and suppresses systematics that may be introduced by the use of photometric background redshifts. We detect the tomographic shear signals at $3.1\sigma$ (A1767) and $3.5\sigma$ (A2065). The shear signal amplitudes are consistent with the cluster dynamical (caustic) masses and they scale appropriately with source redshift. However, comparison with the first spectrophotometric detection of A2029 based on DECam imaging reveals some subtle potential systematic issues in deriving the shear signal for the relatively bright background galaxies used in the analysis. These issues may be important for understanding future more extensive applications of spectrotomography based on further Subaru imaging, as well as Euclid and LSST data. The total of three spectrophotometric detections (A1767, A2029, and A2065) sets the stage for broader application of the technique for unbiased cluster weak lensing mass determinations and potentially for a geometric cosmological test that is independent of other methods.

We perform a cosmological zoom-in simulation of a massive galaxy ($M_s\sim10^{10}\rm M_\odot$ at $z\sim10$) using the GIZMO code. By employing $\leq 3\rm pc$ resolution and a $3.4\rm Myr$ supernova feedback delay, we capture the feedback-free starbursts (FFB) in clusters. The simulation reproduces FFB model predictions and super-bright galaxies observed by JWST. At $z\sim10$, cold streams feed a compact galaxy ($R_{\rm e}\sim1\rm kpc$), with stellar and surface densities ($>10^5\rm cm^{-3}$, $>10^5\rm M_\odot pc^{-2}$) exceeding FFB thresholds. The global star-formation efficiency (SFE) is $\varepsilon_s\sim0.2\text{--}0.3$, associated with a fluctuating star-formation history. We identified over $10^5$ star clusters ($M_{\star}>10^{4.5}\rm M_\odot$) with a nearly scale-free mass distribution (${\rm d}N/{{\rm d}\log M}\propto M^{-1.06}$). Approximately 90\% of star formation occurs in clusters, which at a given time constitute $30\text{--}40\%$ of the total stellar mass. The star formation in most of the clusters of masses $<10^7\rm M_\odot$, occurs in bursts of $<3\rm Myr$ and a local SFE $\sim0.5\pm 0.2$. Cluster metallicities ($-2.01<\log (Z/Z_\odot)<-0.45$) indicate rapid baryon recycling. Feedback-driven outflows exhibit typical temperature of $10^7\rm K$ and typical velocities of $\sim 2000\rm km\ s^{-1}$. In the highly dynamic central $1\rm kpc$, clusters undergo rapid orbital decay and merge to assemble the oblate nuclear stellar cluster. Cluster shapes range from oblate to prolate, with a triaxial median. These clusters are consistent with JWST observations, and a fraction of them may survive to yield the globular clusters (GCs) at low redshifts.

We constructed the 2D model of the Galaxy which initially includes an analytical bar, bulge, disk and halo. The model disk forms the outer elliptical resonance rings R1 and R2 located near the outer Lindblad resonance of the bar (OLR), as well as the inner resonance ring r located near the corotation radius (CR). As the density of stars in the elliptical rings increased, we introduced additional gravitational perturbations created by the rings. The radial component of gravitational perturbations from the elliptical rings, F_R, at a point with the Galactocentric coordinates (R, theta) was represented as a combination of three polynomials in powers R/Re or Re/R, where Re is the distance to the midline (middle) of the ring at a given angle theta. The azimuthal component of the disturbances, F_T, was calculated using the force F_R. The difference between the values of the force F_R (F_T) calculated using the numerical differentiation of the potential and using the analytical representation does not exceed 5.7% (1.3%) of the maximum value of the force F_R generated by the elliptical rings. In general, the gravity of the elliptical rings has little effect on the process of adjustment of epicyclic motions near the OLR of the bar.

The inner Milky Way hosts overlapping stellar populations (bar--bulge, inner thin and thick disks, and halo), complicating population assignments along the line of sight. A joint chemical--dynamical approach is required to isolate a clean field bulge sample. Metal-poor bulge stars are valuable as they likely trace the earliest phases of chemical enrichment in the inner Galaxy. We aim to characterize the alpha-element (Si, Mg) and selected Fe-peak abundances of a dynamically defined sample of bulge field stars, and to contrast these trends with the inner-halo tail in the metal-poor regime. We analyze metal-poor candidates from the Pristine Inner Galaxy Survey observed by the bulge Cluster APOGEE Survey. Using APOGEE/ASPCAP abundances (S/N >= 50), we integrate full 6D orbits in a barred Milky Way potential. Bulge membership is defined via orbital confinement using an apocenter cut and the high-density locus in the (E_J, L_z) plane, with uncertainties estimated using Monte Carlo simulations. We identify 98 stars as bulge members. The metallicity distribution spans -2.5 <= [Fe/H] <= -0.4, with a median [Fe/H] = -1.71, offset to higher metallicity than the inner halo (median [Fe/H] = -1.96). The sample follows a high-alpha sequence with slopes d[Si/Fe]/d[Fe/H] = -0.020(+0.012,-0.051) and d[Mg/Fe]/d[Fe/H] = -0.097(+0.062,-0.133). Fe-peak tracers show [Ni/Fe] ~ 0, while [Mn/Fe] declines with [Fe/H] with a mild upturn at higher metallicity. We detect no significant [Si/Fe] gradients with R_apo or Z_max (+0.010(+0.018,0.000) and +0.006(+0.022,-0.011) dex kpc^-1, respectively). Results are insensitive to Omega_p, yielding indistinguishable memberships. The chemo-orbital evidence favors an in situ origin within the inner Galaxy for the metal-poor bulge field, enriched at early times and rearranged by secular bar evolution, with a minor contribution from halo stars at the most metal-poor end.

Modern surveys such as DESI and \textit{Euclid}, which collect data for hundreds of millions of galaxies to map the large-scale structure (LSS) of the Universe, hold the key to determining the cosmological parameters and testing new physics. This ambition, however, is limited by uncertainties in the galaxy-halo connection: the link between observed galaxies and the underlying, unobservable matter field, by accounting for effects such as galaxy bias and assembly bias (AB). These are particularly poorly-understood for modified gravity (MG) models, which are popular alternatives to the cosmological constant to explain accelerated expansion. We approach this problem using mock emission line galaxy (ELG) and luminous red galaxy (LRG) catalogues in $f(R)$ gravity matching the specifications of ongoing Stage-IV galaxy surveys, generated from state-of-the-art MG hydrodynamical simulations. While the interplay between MG -- especially the chameleon screening mechanism -- and galaxy formation leaves complicated imprints in the galaxy-halo connection, a simple physical picture emerges in which halo and galaxy formation are enhanced for progressively more massive haloes over time. We confirm that the basic galaxy-halo connection model, the halo occupation distribution (HOD), in which galaxy occupation is determined solely by halo mass, underestimates galaxy clustering strength in $\Lambda$CDM by $10$--$20\%$ at $z\lesssim1$ when neglecting AB, and demonstrate that MG introduces further complexity. Extending this model with a suitably-chosen environment density as a secondary HOD variable reduces the AB effect in all models to $2$--$3\%$ for $z\lesssim0.5$. This provides a well-motivated starting point for further works on minimising the impact of AB when testing non-standard cosmological models with LSS.

Accurate modelling of the covariance of clustering observables is essential to fully exploit current and future survey data, which is expected to constrain large-scale clustering signals with unprecedented precision. Computational costs of simulation-based estimates motivate analytical approaches, especially in light of the growing interest towards multi-tracer analyses and parity-odd signatures in two-point statistics, which respectively mitigate cosmic variance and probe relativistic projection effects on cosmological scales. In this work, we generalise previous theoretical results for the Gaussian covariance of multi-tracer power spectrum measurements, providing a general expression applicable to both real (even-parity) and complex (both even- and odd-parity) power spectra. We focus on a generic weighted estimator at first, and then showcase how our general formalism applies to Legendre power spectrum multipoles and two-dimensional power spectrum, recovering known limits in appropriate cases. We validate our predictions against Gaussian Monte Carlo simulations and investigate the structure of the covariance matrix, including the Hermitian properties of its imaginary part.

We derive both the mid-plane and off-plane rotation curves, $v_c(R,z)$, and the vertical acceleration, $a_z(R,z)$, of the Milky Way (MW) using \textit{Gaia}~DR3 data over the ranges of vertical heights $z \in (-2,2)\,$ kpc and galactocentric distances $R \in (8.5,14)$ kpc where the velocity components are determined with high precision, i.e., with an error $< 5\%$. In contrast, the vertical acceleration $a_z(R,z)$ is dominated by model-dependent systematics, with uncertainties of up to $\sim 20\%$. This level of accuracy allows us to place stringent constraints on the geometry of the MW's dark matter (DM) distribution, as the vertical gradients of the gravitational potential attain their maximum within this range of radial and vertical distances corresponding to the characteristic scales of the disk. We find that models including the observed stellar components together with a spherical DM halo fail to reproduce both the pronounced variation of $v_c(R,z)$ with height and the observed behavior of $a_z(R,z)$. In particular, spherical halos with a scale radius of $r_s \sim 15$ kpc contribute negligibly to the off-plane rotation curve and vertical acceleration in the inner disk, leaving these features primarily determined by the stellar mass distribution. Conversely, models in which DM is confined to a flattened, disk-like configuration predict substantial contributions to both $v_c(R,z)$ and $a_z(R,z)$, resulting in a markedly better agreement with the data. We conclude that disk-like DM distributions are strongly favored over spherical halo models. Forthcoming Gaia data releases will enable even more stringent tests of the geometry and distribution of the MW's DM component.

The spectra of an accretion disk around black holes are the basic diagnostic tool to enlighten the underlying flows and then black holes. Accretion flows around black holes, however, are controlled by parameters like the magnetic field, spin of the black hole, accretion rate and temperature of the flow. These quantities affect the (magneto)hydrodynamics of the flow thus consequently lead to variations in the spectrum. We first consider numerical steady state magnetohydrodynamic (MHD) solutions of magnetized accretion flows around black holes to study the dependence of the spectra on these disk properties. The spectrum exhibits strong dependence on the spin of the black hole, accretion rate, magnetic field and the electron temperature of the flow. Variations in these quantities influence the emission peaks and overall luminosity, which can be a tell-tale sign to extract physics of observed spectra. We further validate our results with general relativistic MHD (GRMHD) simulations using the standard and normal evolution (SANE) and magnetically arrested disk (MAD) vector potentials. We consider two black hole spins ($a=0.5$ and $a=0.9375$) to model the magnetic field configurations and study the resulting spectra by comparing MAD and SANE results. We find a large difference in the bolometric luminosities and the location of the emission peaks between SANE and MAD flows. Certain properties of the spectra, like, the ratio of synchrotron radiation to synchrotron self-Comptonization peaks in SANE and MAD, show drastically distinct features. The overall luminosity combined with such metrics can distinguish the magnetic field characteristics in astrophysical systems.