Publications

The sensitivity of X-ray facilities and our ability to detect fainter active galactic nuclei (AGNs) will increase with the upcoming Athena mission and the AXIS and Lynx concept missions, thus improving our understanding of supermassive black holes (BHs) in a luminosity regime that can be dominated by X-ray binaries. We analyze the population of faint AGN (L_x (2-10 keV) 2, while XRBs dominate in some simulations at z<2. Whether the AGN or XRB emission dominates in star-forming and quenched galaxies depends on the simulations. These differences in simulations can be used to discriminate between galaxy formation models with future high-resolution X-ray observations. We compare the luminosity of simulated faint AGN host galaxies to observations of stacked galaxies from Chandra. Our comparison indicates that the simulations post-processed with our X-ray modeling tend to overestimate the AGN+XRB X-ray luminosity; luminosity that can be strongly affected by AGN obscuration. Some simulations reveal clear AGN trends as a function of stellar mass (e.g., galaxy luminosity drop in massive galaxies), which are not apparent in the observations.

We characterize mass, momentum, energy and metal outflow rates of multi-phase galactic winds in a suite of FIRE-2 cosmological "zoom-in" simulations from the Feedback in Realistic Environments (FIRE) project. We analyze simulations of low-mass dwarfs, intermediate-mass dwarfs, Milky Way-mass halos, and high-redshift massive halos. Consistent with previous work, we find that dwarfs eject about 100 times more gas from their interstellar medium (ISM) than they form in stars, while this mass "loading factor" drops below one in massive galaxies. Most of the mass is carried by the hot phase (>105 K) in massive halos and the warm phase (103−105 K) in dwarfs; cold outflows (<103 K) are negligible except in high-redshift dwarfs. Energy, momentum and metal loading factors from the ISM are of order unity in dwarfs and significantly lower in more massive halos. Hot outflows have 2−5× higher specific energy than needed to escape from the gravitational potential of dwarf halos; indeed, in dwarfs, the mass, momentum, and metal outflow rates increase with radius whereas energy is roughly conserved, indicating swept up halo gas. Burst-averaged mass loading factors tend to be larger during more powerful star formation episodes and when the inner halo is not virialized, but we see effectively no trend with the dense ISM gas fraction. We discuss how our results can guide future controlled numerical experiments that aim to elucidate the key parameters governing galactic winds and the resulting associated preventative feedback.

Despite over three hundred years of effort, no solutions exist for predicting when a general planetary configuration will become unstable. We introduce a deep learning architecture to push forward this problem for compact systems. While current machine learning algorithms in this area rely on scientist-derived instability metrics, our new technique learns its own metrics from scratch, enabled by a novel internal structure inspired from dynamics theory. Our Bayesian neural network model can accurately predict not only if, but also when a compact planetary system with three or more planets will go unstable. Our model, trained directly from short N-body time series of raw orbital elements, is more than two orders of magnitude more accurate at predicting instability times than analytical estimators, while also reducing the bias of existing machine learning algorithms by nearly a factor of three. Despite being trained on compact resonant and near-resonant three-planet configurations, the model demonstrates robust generalization to both non-resonant and higher multiplicity configurations, in the latter case outperforming models fit to that specific set of integrations. The model computes instability estimates up to five orders of magnitude faster than a numerical integrator, and unlike previous efforts provides confidence intervals on its predictions. Our inference model is publicly available in the SPOCK package, with training code open-sourced.

We examine the magneto-optical response of chiral multifold fermions. Specifically, we show that they are ideal candidates for observing the Helical Magnetic Effect (HME) previously predicted for simple Weyl fermions. Unlike Weyl fermions, the HME is present in multifold fermions even in the simplest case where the low-energy dispersion is linear and spherically symmetric. In this ideal case, we derive an analytical expression for the HME and find it is proportional to the circular photogalvanic effect; for realistic parameters and accounting for the geometry of the setup, the HME photocurrent could be roughly the same order of magnitude as the circular photogalvanic effect observed in multifold fermions. Additional non-linear and symmetry-breaking terms will ruin the quantization but not hurt the observation of the HME.

Charge control of color centers in semiconductors promises opportunities for novel forms of sensing and quantum information processing. Here, we articulate confocal fluorescence microscopy and magnetic resonance protocols to induce and probe charge transport between discrete sets of engineered nitrogen-vacancy (NV) centers in diamond, down to the level of individual defects. In our experiments, a "source" NV undergoes optically-driven cycles of ionization and recombination to produce a stream of photo-generated carriers, one of which we subsequently capture via a "target" NV several micrometers away. We use a spin-to-charge conversion scheme to encode the spin state of the source color center into the charge state of the target, in the process allowing us to set an upper bound to carrier injection from other background defects. We attribute our observations to the action of unscreened Coulomb potentials producing giant carrier capture cross-sections, orders of magnitude greater than those typically attained in ensemble measurements. Besides their fundamental interest, these results open intriguing prospects in the use of free carriers as a quantum bus to mediate effective interactions between paramagnetic defects in a solid-state chip.

Intentional breaking of the lattice symmetry in solids is a key concept to alter the properties of materials by modifying their electronic band structure. However, the correlation of strain-induced effects and breaking of the lattice symmetry is often indirect, resorting to vibrational spectroscopic techniques such as Raman scattering. Here, we demonstrate that reflectance anisotropy spectroscopy (RAS), which directly depends on the complex dielectric function, enables the direct observation of electronic band structure modulation. Studying the strain-induced symmetry breaking in copper, we show how uniaxial strain lifts the degeneracy of states in the proximity of the both L and X symmetry points, thus altering the matrix element for interband optical transitions, directly observable in RAS. We corroborate our experimental results by analysing the strain-induced changes in the electronic structure based on ab-initio density functional theory calculations. The versatility to study breaking of the lattice symmetry by simple reflectance measurements opens up the possibility to gain a direct insight on the band-structure of other strain-engineered materials, such as graphene and two-dimensional (2D) transition metal dichalcogenides (TMDCs).

Despite over three hundred years of effort, no solutions exist for predicting when a general planetary configuration will become unstable. We introduce a deep learning architecture to push forward this problem for compact systems. While current machine learning algorithms in this area rely on scientist-derived instability metrics, our new technique learns its own metrics from scratch, enabled by a novel internal structure inspired from dynamics theory. Our Bayesian neural network model can accurately predict not only if, but also when a compact planetary system with three or more planets will go unstable. Our model, trained directly from short N-body time series of raw orbital elements, is more than two orders of magnitude more accurate at predicting instability times than analytical estimators, while also reducing the bias of existing machine learning algorithms by nearly a factor of three. Despite being trained on compact resonant and near-resonant three-planet configurations, the model demonstrates robust generalization to both non-resonant and higher multiplicity configurations, in the latter case outperforming models fit to that specific set of integrations. The model computes instability estimates up to five orders of magnitude faster than a numerical integrator, and unlike previous efforts provides confidence intervals on its predictions. Our inference model is publicly available in the SPOCK package, with training code open-sourced.

Indications of disequilibrium throughout the Milky Way (MW) highlight the need for compact, flexible, non-parametric descriptions of phase–space distributions of galaxies. We present a new representation of the current dark matter (DM) distribution and potential derived from N-body simulations of the MW and Large Magellanic Cloud (LMC) system using basis function expansions (BFEs). We incorporate methods to maximize the physical signal in the representation. As a result, the simulations of 108 DM particles representing the distorted MW(MW+LMC) system can be described by ∼236(2067) coefficients. We find that the LMC induces asymmetric perturbations (odd l, m) to the MW’s halo, which are inconsistent with oblate, prolate, or triaxial halos. Furthermore, the energy in high order even modes (l, m > 2) is similar to average triaxial halos found in cosmological simulations. As such, the response of the MW’s halo to the LMC must be accounted for in order to recover the imprints of its assembly history. The LMC causes the outer halo (>30 kpc) to shift from the disk center of mass (COM) by ∼15–25 kpc at present day, manifesting as a dipole in the BFE and in the radial velocities of halo stars. The shift depends on the LMC’s infall mass, the distortion of the LMC’s halo and the MW halo response.Within 30 kpc, halo tracers are expected to orbit the COM of the MW’s disk, regardless of LMC infall mass. The LMC’s halo is also distorted by MW tides; we discuss the implications for its mass loss and the subsequent effects on current Magellanic satellites.

Stacking is a widely used model averaging technique that asymptotically yields optimal predictions among linear averages. We show that stacking is most effective when model predictive performance is heterogeneous in inputs, and we can further improve the stacked mixture with a hierarchical model. We generalize stacking to Bayesian hierarchical stacking. The model weights are varying as a function of data, partially-pooled, and inferred using Bayesian inference. We further incorporate discrete and continuous inputs, other structured priors, and time series and longitudinal data. To verify the performance gain of the proposed method, we derive theory bounds, and demonstrate on several applied problems.