Aframe neural network achieves matched-filter sensitivity for binary neutron star GW searches at lower computational cost using heterodyning and a single GPU.
Fast Fisher Matrices and Lazy Likelihoods
13 Pith papers cite this work. Polarity classification is still indexing.
13
Pith papers citing it
abstract
Theoretical studies in gravitational wave astronomy often require the calculation of Fisher Information Matrices and Likelihood functions, which in a direct approach entail the costly step of computing gravitational waveforms. Here I describe an alternative technique that sidesteps the need to compute full waveforms, resulting in significant computational savings. I describe how related techniques can be used to speed up Bayesian inference applied to real gravitational wave data.
citation-role summary
background 3
method 1
citation-polarity summary
representative citing papers
Ratio-Filter Dechirping converts gravitational-wave matched filtering from a memory-bound FFT into a cache-efficient FIR convolution, delivering a measured 8x speedup in the core loop.
For a loud stellar-mass black hole binary, adding realistic orbital eccentricity to multiband Bayesian analysis weakens dipole-radiation constraints to |b|≲10⁻⁷, with strong degeneracies among dipole strength, chirp mass, and eccentricity.
Direct sampling under a uniform-in-dL prior raises P(H0>120) from 0.017 to 0.159 for GW170817; post-hoc reweighting recovers only 0.041 because a low-dL mode is undersampled.
Simulations show a 40-50 solar-mass black-hole cutoff is not guaranteed to be confidently recovered from GWTC-4-like catalogs, spurious detections are unlikely, and O4 data would reduce cutoff-mass uncertainty by at least 20 percent while yielding only a lower bound on the carbon-alpha reaction rate
Forecasts that golden and silver dark sirens with HETDEX VIRUS follow-up can constrain H0 to a few percent using one year of LIGO-A# observations for z < 0.2 events.
Neural posterior estimation shows a two-misaligned-L ET network yields better sky and volume localization for high-redshift massive BBHs than triangular ET, with fewer multimodalities, though luminosity distance estimates are less precise.
Neural network surrogate approximates precessing compact binary gravitational waveforms up to 1000x faster than the base EOB model with validated accuracy.
Domain-optimized neural posterior estimation (labrador) reaches ~1% median importance-sampling efficiency on aligned-spin GW signals and covers long-duration systems with m2 < 10 M☉.
Bilby-antiglitch jointly models astrophysical signals and quasi-physical glitches to recover true source properties from simulated gravitational wave data contaminated by loud non-Gaussian transients.
ET-Δ outperforms ET-2L on luminosity distance and source-frame masses for BBH events because its redundancy produces higher multi-detector uptime under realistic duty-cycle modeling.
Simulations of continuous-wave searches show that PTA data first constrain GW frequency and strain amplitude together, then sky location, with chirp mass and inclination following later for evolving sources, with precision depending on source frequency and sky position.
BILBY is validated on simulated compact binary signals and reproduces the eleven GWTC-1 results with configuration and output files provided for reproduction.
citing papers explorer
-
AI-enabled gravitational-waves searches for binary neutron stars at optimal sensitivity
Aframe neural network achieves matched-filter sensitivity for binary neutron star GW searches at lower computational cost using heterodyning and a single GPU.
-
Beyond FINDCHIRP: Breaking the memory wall and optimal FFTs for Gravitational-Wave Matched-Filter Searches with Ratio-Filter Dechirping
Ratio-Filter Dechirping converts gravitational-wave matched filtering from a memory-bound FFT into a cache-efficient FIR convolution, delivering a measured 8x speedup in the core loop.
-
Constraining Dipole Radiation with Multiband Gravitational Waves from Eccentric Binary Black Holes
For a loud stellar-mass black hole binary, adding realistic orbital eccentricity to multiband Bayesian analysis weakens dipole-radiation constraints to |b|≲10⁻⁷, with strong degeneracies among dipole strength, chirp mass, and eccentricity.
-
Rapid Hubble constant inference from GW170817 using GPU-accelerated nested sampling: prior sensitivity and the limits of post-hoc reweighting
Direct sampling under a uniform-in-dL prior raises P(H0>120) from 0.017 to 0.159 for GW170817; post-hoc reweighting recovers only 0.041 because a low-dL mode is undersampled.
-
Measurement prospects for the pair-instability mass cutoff with gravitational waves
Simulations show a 40-50 solar-mass black-hole cutoff is not guaranteed to be confidently recovered from GWTC-4-like catalogs, spurious detections are unlikely, and O4 data would reduce cutoff-mass uncertainty by at least 20 percent while yielding only a lower bound on the carbon-alpha reaction rate
-
Golden and Silver Dark Sirens for precise H0 measurement with HETDEX
Forecasts that golden and silver dark sirens with HETDEX VIRUS follow-up can constrain H0 to a few percent using one year of LIGO-A# observations for z < 0.2 events.
-
Comparing next-generation detector configurations for high-redshift gravitational wave sources with neural posterior estimation
Neural posterior estimation shows a two-misaligned-L ET network yields better sky and volume localization for high-redshift massive BBHs than triangular ET, with fewer multimodalities, though luminosity distance estimates are less precise.
-
Fast neural network surrogate for multimodal effective-one-body gravitational waveforms from generically precessing compact binaries
Neural network surrogate approximates precessing compact binary gravitational waveforms up to 1000x faster than the base EOB model with validated accuracy.
-
labrador: A domain-optimized machine-learning tool for gravitational wave inference
Domain-optimized neural posterior estimation (labrador) reaches ~1% median importance-sampling efficiency on aligned-spin GW signals and covers long-duration systems with m2 < 10 M☉.
-
A parametric signal plus noise inference framework for short duration non-Gaussian noise transients
Bilby-antiglitch jointly models astrophysical signals and quasi-physical glitches to recover true source properties from simulated gravitational wave data contaminated by loud non-Gaussian transients.
-
Impact of the Einstein Telescope's duty cycle on the estimation of binary black holes parameters
ET-Δ outperforms ET-2L on luminosity distance and source-frame masses for BBH events because its redundancy produces higher multi-detector uptime under realistic duty-cycle modeling.
-
Expectations for the first supermassive black-hole binary resolved by PTAs II: Milestones for binary characterization
Simulations of continuous-wave searches show that PTA data first constrain GW frequency and strain amplitude together, then sky location, with chirp mass and inclination following later for evolving sources, with precision depending on source frequency and sky position.
-
Bayesian inference for compact binary coalescences with BILBY: Validation and application to the first LIGO--Virgo gravitational-wave transient catalogue
BILBY is validated on simulated compact binary signals and reproduces the eleven GWTC-1 results with configuration and output files provided for reproduction.