Search

Generating high-resolution simulations is key for advancing our understanding of one of the universe's most violent events: Black Hole mergers. However, generating Black Hole simulations is limited by prohibitive computational costs and scalability issues, reducing the simulation's fidelity and resolution achievable within reasonable time frames and resources. In this work, we introduce a novel method that circumvents these limitations by applying a super-resolution technique without directly needing high-resolution labels, leveraging the Hamiltonian and momentum constraints-fundamental equations in general relativity that govern the dynamics of spacetime. We demonstrate that our method achieves a reduction in constraint violation by one to two orders of magnitude and generalizes effectively to out-of-distribution simulations., Comment: Code available at https://github.com/ThomasHelfer/TorchGRTL and data at https://huggingface.co/datasets/thelfer/BinaryBlackHole

Scaling laws for large language models (LLMs) have provided useful guidance on how to train ever larger models for predictable performance gains. Time series forecasting shares a similar sequential structure to language, and is amenable to large-scale transformer architectures. Here we show that foundational decoder-only time series transformer models exhibit analogous scaling-behavior to LLMs, while architectural details (aspect ratio and number of heads) have a minimal effect over broad ranges. We assemble a large corpus of heterogenous time series data on which to train, and establish, for the first time, power-law scaling relations with respect to parameter count, dataset size, and training compute, spanning five orders of magnitude., Comment: 8 pages, 3 figures

We present a framework for the efficient computation of optimal Bayesian decisions under intractable likelihoods, by learning a surrogate model for the expected utility (or its distribution) as a function of the action and data spaces. We leverage recent advances in simulation-based inference and Bayesian optimization to develop active learning schemes to choose where in parameter and action spaces to simulate. This allows us to learn the optimal action in as few simulations as possible. The resulting framework is extremely simulation efficient, typically requiring fewer model calls than the associated posterior inference task alone, and a factor of $100-1000$ more efficient than Monte-Carlo based methods. Our framework opens up new capabilities for performing Bayesian decision making, particularly in the previously challenging regime where likelihoods are intractable, and simulations expensive., Comment: 12 pages, 4 figures

We consider state and parameter estimation for compartmental models having both time-varying and time-invariant parameters. Though the described Bayesian computational framework is general, we look at a specific application to the susceptible-infectious-removed (SIR) model which describes a basic mechanism for the spread of infectious diseases through a system of coupled nonlinear differential equations. The SIR model consists of three states, namely, the three compartments, and two parameters which control the coupling among the states. The deterministic SIR model with time-invariant parameters has shown to be overly simplistic for modelling the complex long-term dynamics of diseases transmission. Recognizing that certain model parameters will naturally vary in time due to seasonal trends, non-pharmaceutical interventions, and other random effects, the estimation procedure must systematically permit these time-varying effects to be captured, without unduly introducing artificial dynamics into the system. To this end, we leverage the robustness of the Markov Chain Monte Carlo (MCMC) algorithm for the estimation of time-invariant parameters alongside nonlinear filters for the joint estimation of the system state and time-varying parameters. We demonstrate performance of the framework by first considering a series of examples using synthetic data, followed by an exposition on public health data collected in the province of Ontario.

Ranked data is commonly used in research across many fields of study including medicine, biology, psychology, and economics. One common statistic used for analyzing ranked data is Kendall's {\tau} coefficient, a non-parametric measure of rank correlation which describes the strength of the association between two monotonic continuous or ordinal variables. While the mathematics involved in calculating Kendall's {\tau} is well-established, there are relatively few graphing methods available to visualize the results. Here, we describe a visualization method and provide an interactive app for graphing Kendall's {\tau} which uses a series of rigid Euclidean transformations along a Cartesian plane to map rank pairs into discrete quadrants. The resulting graph provides a visualization of rank correlation which helps display the proportion of concordant and discordant pairs. Moreover, this method highlights other key features of the data which are not represented by Kendall's {\tau} alone but may nevertheless be meaningful, such as the relationship between discrete pairs of observations. We demonstrate the effectiveness of our approach through several examples and compare our results to other visualization methods.

We investigate the possibility of improving the accuracy of the phenomenological waveform model, IMRPhenomD, by jointly optimizing all the calibration coefficients at once, given a set of numerical relativity (NR) waveforms. When IMRPhenomD was first calibrated to NR waveforms, different parts (i.e., the inspiral, merger, and ringdown) of the waveform were calibrated separately. Using ripple, a library of waveform models compatible with automatic differentiation, we can, for the first time, perform gradient-based optimization on all the waveform coefficients at the same time. This joint optimization process allows us to capture previously ignored correlations between separate parts of the waveform. We found that after recalibration, the median mismatch between the model and NR waveforms decreases by 50%. We further explore how different regions of the source parameter space respond to the optimization procedure. We find that the degree of improvement correlates with the spins of the source. This work shows a promising avenue to help understand and treat systematic error in waveform models., Comment: 11 pages, 9 figures

We report results on the first matched-filtering search for binaries with compact objects having large tidal deformabilities in the LIGO-Virgo gravitational wave (GW) data. The tidal deformability of a body is quantified by the ``Love number" $\Lambda \propto \hskip 1pt (r/m)^5$, where $r/m$ is the body's (inverse) compactness. Due to its strong dependence on compactness, the $\Lambda$ of larger-sized compact objects can easily be many orders of magnitude greater than those of black holes and neutron stars, leaving phase shifts which are sufficiently large for these binaries to be missed by binary black hole (BBH) templated searches. In this paper, we conduct a search using inspiral-only waveforms with zero spins but finite tides, with the search space covering chirp masses $3 M_\odot < \mathcal{M} < 15 M_\odot$ and effective tidal deformabilities $10^2 \lesssim \tilde{\Lambda} \lesssim 10^6$. We find no statistically significant GW candidates. This null detection implies an upper limit on the merger rate of such binaries in the range $[1-300] \hskip 2pt \text{Gpc}^{-3} \text{year}^{-1}$, depending on $\mathcal{M}$ and $\tilde{\Lambda}$. While our constraints are model agnostic, we discuss the implications on beyond the Standard Model scenarios that give rise to boson stars and superradiant clouds. Using inspiral-only waveforms we recover many of the BBH signals which were previously identified with full inspiral-merger-ringdown templates. We also constrain the Love number of black holes to $\Lambda \lesssim 10^3$ at the 90\% credible interval. Our work is the first-ever dedicated template-based search for compact objects that are not only black holes and neutron stars. Additionally, our work demonstrates a novel way of finding new physics in GW data, widening the scope of potential discovery to previously unexplored parameter space., Comment: 42+11 pages, 12 figures

We present a lightweight, flexible, and high-performance framework for inferring the properties of gravitational-wave events. By combining likelihood heterodyning, automatically-differentiable and accelerator-compatible waveforms, and gradient-based Markov chain Monte Carlo (MCMC) sampling enhanced by normalizing flows, we achieve full Bayesian parameter estimation for real events like GW150914 and GW170817 within a minute of sampling time. Our framework does not require pretraining or explicit reparameterizations and can be generalized to handle higher dimensional problems. We present the details of our implementation and discuss trade-offs and future developments in the context of other proposed strategies for real-time parameter estimation. Our code for running the analysis is publicly available on GitHub https://github.com/kazewong/jim.

We propose the use of automatic differentiation through the programming framework jax for accelerating a variety of analysis tasks throughout gravitational wave (GW) science. Firstly, we demonstrate that complete waveforms which cover the inspiral, merger, and ringdown of binary black holes (i.e. IMRPhenomD) can be written in jax and demonstrate that the serial evaluation speed of the waveform (and its derivative) is similar to the lalsuite implementation in C. Moreover, jax allows for GPU-accelerated waveform calls which can be over an order of magnitude faster than serial evaluation on a CPU. We then focus on three applications where efficient and differentiable waveforms are essential. Firstly, we demonstrate how gradient descent can be used to optimize the $\sim 200$ coefficients that are used to calibrate the waveform model. In particular, we demonstrate that the typical match with numerical relativity waveforms can be improved by more than 50% without any additional overhead. Secondly, we show that Fisher forecasting calculations can be sped up by $\sim 100\times$ (on a CPU) with no loss in accuracy. This increased speed makes population forecasting substantially simpler. Finally, we show that gradient-based samplers like Hamiltonian Monte Carlo lead to significantly reduced autocorrelation values when compared to traditional Monte Carlo methods. Since differentiable waveforms have substantial advantages for a variety of tasks throughout GW science, we propose that waveform developers use jax to build new waveforms moving forward. Our waveform code, ripple, can be found at https://github.com/tedwards2412/ripple, and will continue to be updated with new waveforms as they are implemented., Comment: 12 pages, 6 figures, code can be found at https://github.com/tedwards2412/ripple

We present highly accurate, dimensionally-reduced gravitational waveforms for binary inspirals whose components have large spin-induced quadrupole moments. The spin-induced quadrupole of a body first appears in the phase of a waveform at the early inspiral stage of the binary coalescence, making it a relatively clean probe of the internal structure of the body. However, for objects with large quadrupolar deviations from Kerr, searches using binary black hole (BBH) models would be ineffective. In order to perform a computationally-feasible search, we present two dimensionally-reduced models which are derived from the original six-dimensional post-Newtonian waveform for such systems. Our dimensional reduction method is guided by power counting in the post-Newtonian expansion, suitable reparameterizations of the source physics, and truncating terms in the phase that are small in most physically well-motivated regions of parameter space. In addition, we note that large quadrupolar deviations cause the frequency at which a binary system reaches its minimum binding energy to be reduced substantially. This minimum signals the end of the inspiral regime and provides a natural cutoff for the PN waveform. We provide accurate analytic estimates for these frequency cutoffs. Finally, we perform injection studies to test the effectualness of the dimensionally reduced waveforms. We find that over $80\%$ of the injections have an effectualness of $\varepsilon > 0.999$, significantly higher than is typically required for standard BBH banks, for systems with component spins of $|\chi_i| \lesssim 0.6$ and dimensionless quadrupole of $\kappa_i \lesssim 10^3$. Importantly, these waveforms represent an essential first step towards enabling an effective search for astrophysical objects with large quadrupoles., Comment: 18 pages, 10 figures

A nonlinear partial differential equation (PDE) based compartmental model of COVID-19 provides a continuous trace of infection over space and time. Finer resolutions in the spatial discretization, the inclusion of additional model compartments and model stratifications based on clinically relevant categories contribute to an increase in the number of unknowns to the order of millions. We adopt a parallel scalable solver allowing faster solutions for these high fidelity models. The solver combines domain decomposition and algebraic multigrid preconditioners at multiple levels to achieve the desired strong and weak scalability. As a numerical illustration of this general methodology, a five-compartment susceptible-exposed-infected-recovered-deceased (SEIRD) model of COVID-19 is used to demonstrate the scalability and effectiveness of the proposed solver for a large geographical domain (Southern Ontario). It is possible to predict the infections up to three months for a system size of 92 million (using 1780 processes) within 7 hours saving months of computational effort needed for the conventional solvers.