Publications

Upcoming measurements of the small-scale primary cosmic microwave background (CMB) temperature and polarization power spectra (TT/TE/EE) are anticipated to yield transformative constraints on new physics, including the effective number of relativistic species in the early universe (Neff). However, at multipoles ℓ≳3000, the primary CMB power spectra receive significant contributions from gravitational lensing. While these modes still carry primordial information, their theoretical modeling requires knowledge of the CMB lensing convergence power spectrum, CκκL, including on small scales where it is affected by nonlinear gravitational evolution and baryonic feedback processes. Thus, the high-ℓ primary CMB is sensitive to these late-time, nonlinear effects. Here, we show that inaccuracies in the modeling of CκκL can yield surprisingly large biases on cosmological parameters inferred from the primary CMB power spectra measured by the upcoming Simons Observatory and CMB-S4 experiments. For CMB-S4, the biases can be as large as 1.6σ on the Hubble constant H0 in a fit to ΛCDM and 1.2σ on Neff in a fit to ΛCDM+Neff.
We show that these biases can be mitigated by explicitly discarding all TT data at ℓ>3000 or by marginalizing over parameters describing baryonic feedback processes, both at the cost of slightly larger error bars. We also discuss an alternative, data-driven mitigation strategy based on delensing the CMB T and E-mode maps. Finally, we show that analyses of upcoming data will require Einstein-Boltzmann codes to be run with much higher numerical precision settings than is currently standard, so as to avoid similar -- or larger -- parameter biases due to inaccurate theoretical predictions.

The brain must extract behaviorally relevant latent variables from the signals streamed by the sensory organs. Such latent variables are often encoded in the dynamics that generated the signal rather than in the specific realization of the waveform. Therefore, one problem faced by the brain is to segment time series based on underlying dynamics. We present two algorithms for performing this segmentation task that are biologically plausible, which we define as acting in a streaming setting and all learning rules being local. One algorithm is model based and can be derived from an optimization problem involving a mixture of autoregressive processes. This algorithm relies on feedback in the form of a prediction error and can also be used for forecasting future samples. In some brain regions, such as the retina, the feedback connections necessary to use the prediction error for learning are absent. For this case, we propose a second, model-free algorithm that uses a running estimate of the autocorrelation structure of the signal to perform the segmentation. We show that both algorithms do well when tasked with segmenting signals drawn from autoregressive models with piecewise-constant parameters. In particular, the segmentation accuracy is similar to that obtained from oracle-like methods in which the ground-truth parameters of the autoregressive models are known. We also test our methods on data sets generated by alternating snippets of voice recordings. We provide implementations of our algorithms at https://github.com/ttesileanu/bio-time-series.

To understand the formation and evolution of the Milky Way disk, we must connect its current properties to its past. We explore hydrodynamical cosmological simulations to investigate how the chemical abundances of stars might be linked to their origins. Using hierarchical clustering of abundance measurements in two Milky Way–like simulations with distributed and steady star formation histories, we find that groups of chemically similar stars comprise different groups in birth place (Rbirth) and time (age). Simulating observational abundance errors (0.05 dex), we find that to trace distinct groups of (Rbirth, age) requires a large vector of abundances. Using 15 element abundances (Fe, O, Mg, S, Si, C, P, Mn, Ne, Al, N, V, Ba, Cr, Co), up to ≈10 groups can be defined with ≈25 percent overlap in (Rbirth, age). We build a simple model to show that in the context of these simulations, it is possible to infer a star's age and Rbirth from abundances with precisions of ±0.06 Gyr and ±1.17 kpc, respectively. We find that abundance clustering is ineffective for a third simulation, where low-α stars form distributed in the disk and early high-α stars form more rapidly in clumps that sink toward the Galactic center as their constituent stars evolve to enrich the interstellar medium. However, this formation path leads to large age dispersions across the [α/Fe]–[Fe/H] plane, which is inconsistent with the Milky Way's observed properties. We conclude that abundance clustering is a promising approach toward charting the history of our Galaxy.

The concurrent growth of supermassive black holes (SMBHs) and their host galaxies remains to be fully explored, especially at high redshift. While often understood as a consequence of self-regulation via AGN feedback, it can also be explained by alternative SMBH accretion models. Here, we expand on previous work by studying the growth of SMBHs with the help of a large suite of cosmological zoom-in simulations (MassiveFIRE) that are part of the Feedback in Realistic Environments (FIRE) project. The growth of SMBHs is modelled in post-processing with different black hole accretion models, placements, and merger treatments, and validated by comparing to on-the-fly calculations. Scaling relations predicted by the gravitational torque driven accretion (GTDA) model agree with observations at low redshift without the need for AGN feedback, in contrast to models in which the accretion rate depends strongly on SMBH mass. At high redshift, we find deviations from the local scaling relations in line with previous theoretical results. In particular, SMBHs are under-massive, presumably due to stellar feedback, but start to grow efficiently once their host galaxies reach M∗∼1010M⊙. We analyse and explain these findings in the context of a simple analytic model. Finally, we show that the predicted scaling relations depend sensitively on the SMBH location and the efficiency of SMBH merging, particularly in low-mass systems. These findings highlight the relevance of understanding the evolution of SMBH-galaxy scaling relations to predict the rate of gravitational wave signals from SMBH mergers across cosmic history.

Stars born on near-circular orbits in spiral galaxies can subsequently migrate to different orbits due to interactions with non-axisymmetric disturbances within the disc such as bars or spiral arms. This paper extends the study of migration to examine the role of external influences using the example of the interaction of the Sagittarius dwarf galaxy (Sgr) with the Milky Way (MW). We first make impulse approximation estimates to characterize the influence of Sgr disc passages. The tidal forcing from Sgr can produce changes in both guiding radius (ΔRg) and orbital eccentricity, as quantified by the maximum radial excursion, ΔRmax. These changes follow a quadrupole-like pattern across the face of the disc, with amplitude increasing with Galactocentric radius. We next examine a collisionless N-body simulation of a Sgr-like satellite interacting with a MW-like galaxy and find that Sgr's influence in the outer disc dominates over the secular evolution of orbits between disc passages. Finally, we use the same simulation to explore possible observable signatures of Sgr-induced migration by painting the simulation with different age stellar populations. We find that following Sgr disc passages, the migration it induces manifests within an annulus as an approximate quadrupole in azimuthal metallicity variations (δ[Fe/H]), along with systematic variations in orbital eccentricity, ΔRmax. These systematic variations can persist for several rotational periods. We conclude that this combination of signatures may be used to distinguish between the different migration mechanisms shaping the chemical abundance patterns of the Milky Way's thin disc.

The communities who develop and support open source scientific software packages are crucial to the utility and success of such packages. Moreover, these communities form an important part of the human infrastructure that enables scientific progress. This paper discusses aspects of the PETSc (Portable Extensible Toolkit for Scientific Computation) community, its organization, and technical approaches that enable community members to help each other efficiently

Dense assemblies of self-propelling rods (SPRs) may exhibit fascinating collective behaviors and anomalous physical properties that are far away from equilibrium. Using large-scale Brownian dynamics simulations, we investigate the dynamics of disclination defects in 2D fluidized swarming motions of dense dry SPRs (i.e., without hydrodynamic effects) that form notable local positional topological structures that are reminiscent of smectic order. We find the deformations of smectic-like rod layers can create unique polar structures that lead to slow translations and rotations of ±1/2-order defects, which are fundamentally different from the fast streaming defect motions observed in wet active matter. We measure and characterize the statistical properties of topological defects and reveal their connections with the coherent structures. Furthermore, we construct a bottom-up active-liquid-crystal model to analyze the instability of polar lanes, which effectively leads to defect formation between interlocked polar lanes and serves as the origin of the large-scale swarming motions.

This paper introduces a Bayesian multilevel model based on the age-period-cohort framework to examine Chinese happiness. Using 8 waves of the Chinese General Social Survey (CGSS) data between 2005–2015, the model not only solves the co-linearity problem with weakly informative priors and explicit assumptions, it also produces more computationally stable results. Our estimation results show how Chinese happiness changes in an individual’s life circle and how one’s life experience is accumulated to her/his happiness with cognitive development. We identify some different generation patterns and explain generation differences in happiness across the various birth years with narratives of historical events. This paper contributes to existing studies both theoretically and methodologically. The novel modeling strategy and the analytical framework which assisted with historical narratives altogether explain better the age, period, and cohort effects on Chinese happiness.

Most existing theoretical analysis of reinforcement learning (RL) is limited to the tabular setting or linear models due to the difficulty in dealing with function approximation in high dimensional space with an uncertain environment. This work offers a fresh perspective into this challenge by analyzing RL in a general reproducing kernel Hilbert space (RKHS). We consider a family of Markov decision processes $\mathcal{M}$ of which the reward functions lie in the unit ball of an RKHS and transition probabilities lie in a given arbitrary set. We define a quantity called perturbational complexity by distribution mismatch $\Delta_{\mathcal{M}}(\epsilon)$ to characterize the complexity of the admissible state-action distribution space in response to a perturbation in the RKHS with scale $\epsilon$. We show that $\Delta_{\mathcal{M}}(\epsilon)$ gives both the lower bound of the error of all possible algorithms and the upper bound of two specific algorithms (fitted reward and fitted Q-iteration) for the RL problem. Hence, the decay of $\Delta_\mathcal{M}(\epsilon)$ with respect to $\epsilon$ measures the difficulty of the RL problem on $\mathcal{M}$. We further provide some concrete examples and discuss whether $\Delta_{\mathcal{M}}(\epsilon)$ decays fast or not in these examples. As a byproduct, we show that when the reward functions lie in a high dimensional RKHS, even if the transition probability is known and the action space is finite, it is still possible for RL problems to suffer from the curse of dimensionality.