{% raw %}
Title: Create a Markdown Blog Post Integrating Research Details and a Featured Paper
====================================================================================
This task involves generating a Markdown file (ready for a GitHub-served Jekyll site) that integrates our research details with a featured research paper. The output must follow the exact format and conventions described below.
====================================================================================
Output Format (Markdown):
------------------------------------------------------------------------------------
---
layout: post
title: "Parallel Nested Slice Sampling for Gravitational Wave Parameter Estimation"
date: 2025-09-29
categories: papers
---
Content generated by [gemini-2.5-pro](https://deepmind.google/technologies/gemini/) using [this prompt](/prompts/content/2025-09-29-2509.24949.txt).
Image generated by [imagen-4.0-generate-001](https://deepmind.google/technologies/gemini/) using [this prompt](/prompts/images/2025-09-29-2509.24949.txt).
------------------------------------------------------------------------------------
====================================================================================
Please adhere strictly to the following instructions:
====================================================================================
Section 1: Content Creation Instructions
====================================================================================
1. **Generate the Page Body:**
- Write a well-composed, engaging narrative that is suitable for a scholarly audience interested in advanced AI and astrophysics.
- Ensure the narrative is original and reflective of the tone and style and content in the "Homepage Content" block (provided below), but do not reuse its content.
- Use bullet points, subheadings, or other formatting to enhance readability.
2. **Highlight Key Research Details:**
- Emphasize the contributions and impact of the paper, focusing on its methodology, significance, and context within current research.
- Specifically highlight the lead author ({'name': 'David Yallup'}). When referencing any author, use Markdown links from the Author Information block (choose academic or GitHub links over social media).
3. **Integrate Data from Multiple Sources:**
- Seamlessly weave information from the following:
- **Paper Metadata (YAML):** Essential details including the title and authors.
- **Paper Source (TeX):** Technical content from the paper.
- **Bibliographic Information (bbl):** Extract bibliographic references.
- **Author Information (YAML):** Profile details for constructing Markdown links.
- Merge insights from the Paper Metadata, TeX source, Bibliographic Information, and Author Information blocks into a coherent narrative—do not treat these as separate or isolated pieces.
- Insert the generated narrative between the HTML comments:
and
4. **Generate Bibliographic References:**
- Review the Bibliographic Information block carefully.
- For each reference that includes a DOI or arXiv identifier:
- For DOIs, generate a link formatted as:
[10.1234/xyz](https://doi.org/10.1234/xyz)
- For arXiv entries, generate a link formatted as:
[2103.12345](https://arxiv.org/abs/2103.12345)
- **Important:** Do not use any LaTeX citation commands (e.g., `\cite{...}`). Every reference must be rendered directly as a Markdown link. For example, instead of `\cite{mycitation}`, output `[mycitation](https://doi.org/mycitation)`
- **Incorrect:** `\cite{10.1234/xyz}`
- **Correct:** `[10.1234/xyz](https://doi.org/10.1234/xyz)`
- Ensure that at least three (3) of the most relevant references are naturally integrated into the narrative.
- Ensure that the link to the Featured paper [2509.24949](https://arxiv.org/abs/2509.24949) is included in the first sentence.
5. **Final Formatting Requirements:**
- The output must be plain Markdown; do not wrap it in Markdown code fences.
- Preserve the YAML front matter exactly as provided.
====================================================================================
Section 2: Provided Data for Integration
====================================================================================
1. **Homepage Content (Tone and Style Reference):**
```markdown
---
layout: home
---
The Handley Research Group stands at the forefront of cosmological exploration, pioneering novel approaches that fuse fundamental physics with the transformative power of artificial intelligence. We are a dynamic team of researchers, including PhD students, postdoctoral fellows, and project students, based at the University of Cambridge. Our mission is to unravel the mysteries of the Universe, from its earliest moments to its present-day structure and ultimate fate. We tackle fundamental questions in cosmology and astrophysics, with a particular focus on leveraging advanced Bayesian statistical methods and AI to push the frontiers of scientific discovery. Our research spans a wide array of topics, including the [primordial Universe](https://arxiv.org/abs/1907.08524), [inflation](https://arxiv.org/abs/1807.06211), the nature of [dark energy](https://arxiv.org/abs/2503.08658) and [dark matter](https://arxiv.org/abs/2405.17548), [21-cm cosmology](https://arxiv.org/abs/2210.07409), the [Cosmic Microwave Background (CMB)](https://arxiv.org/abs/1807.06209), and [gravitational wave astrophysics](https://arxiv.org/abs/2411.17663).
### Our Research Approach: Innovation at the Intersection of Physics and AI
At The Handley Research Group, we develop and apply cutting-edge computational techniques to analyze complex astronomical datasets. Our work is characterized by a deep commitment to principled [Bayesian inference](https://arxiv.org/abs/2205.15570) and the innovative application of [artificial intelligence (AI) and machine learning (ML)](https://arxiv.org/abs/2504.10230).
**Key Research Themes:**
* **Cosmology:** We investigate the early Universe, including [quantum initial conditions for inflation](https://arxiv.org/abs/2002.07042) and the generation of [primordial power spectra](https://arxiv.org/abs/2112.07547). We explore the enigmatic nature of [dark energy, using methods like non-parametric reconstructions](https://arxiv.org/abs/2503.08658), and search for new insights into [dark matter](https://arxiv.org/abs/2405.17548). A significant portion of our efforts is dedicated to [21-cm cosmology](https://arxiv.org/abs/2104.04336), aiming to detect faint signals from the Cosmic Dawn and the Epoch of Reionization.
* **Gravitational Wave Astrophysics:** We develop methods for [analyzing gravitational wave signals](https://arxiv.org/abs/2411.17663), extracting information about extreme astrophysical events and fundamental physics.
* **Bayesian Methods & AI for Physical Sciences:** A core component of our research is the development of novel statistical and AI-driven methodologies. This includes advancing [nested sampling techniques](https://arxiv.org/abs/1506.00171) (e.g., [PolyChord](https://arxiv.org/abs/1506.00171), [dynamic nested sampling](https://arxiv.org/abs/1704.03459), and [accelerated nested sampling with $\beta$-flows](https://arxiv.org/abs/2411.17663)), creating powerful [simulation-based inference (SBI) frameworks](https://arxiv.org/abs/2504.10230), and employing [machine learning for tasks such as radiometer calibration](https://arxiv.org/abs/2504.16791), [cosmological emulation](https://arxiv.org/abs/2503.13263), and [mitigating radio frequency interference](https://arxiv.org/abs/2211.15448). We also explore the potential of [foundation models for scientific discovery](https://arxiv.org/abs/2401.00096).
**Technical Contributions:**
Our group has a strong track record of developing widely-used scientific software. Notable examples include:
* [**PolyChord**](https://arxiv.org/abs/1506.00171): A next-generation nested sampling algorithm for Bayesian computation.
* [**anesthetic**](https://arxiv.org/abs/1905.04768): A Python package for processing and visualizing nested sampling runs.
* [**GLOBALEMU**](https://arxiv.org/abs/2104.04336): An emulator for the sky-averaged 21-cm signal.
* [**maxsmooth**](https://arxiv.org/abs/2007.14970): A tool for rapid maximally smooth function fitting.
* [**margarine**](https://arxiv.org/abs/2205.12841): For marginal Bayesian statistics using normalizing flows and KDEs.
* [**fgivenx**](https://arxiv.org/abs/1908.01711): A package for functional posterior plotting.
* [**nestcheck**](https://arxiv.org/abs/1804.06406): Diagnostic tests for nested sampling calculations.
### Impact and Discoveries
Our research has led to significant advancements in cosmological data analysis and yielded new insights into the Universe. Key achievements include:
* Pioneering the development and application of advanced Bayesian inference tools, such as [PolyChord](https://arxiv.org/abs/1506.00171), which has become a cornerstone for cosmological parameter estimation and model comparison globally.
* Making significant contributions to the analysis of major cosmological datasets, including the [Planck mission](https://arxiv.org/abs/1807.06209), providing some of the tightest constraints on cosmological parameters and models of [inflation](https://arxiv.org/abs/1807.06211).
* Developing novel AI-driven approaches for astrophysical challenges, such as using [machine learning for radiometer calibration in 21-cm experiments](https://arxiv.org/abs/2504.16791) and [simulation-based inference for extracting cosmological information from galaxy clusters](https://arxiv.org/abs/2504.10230).
* Probing the nature of dark energy through innovative [non-parametric reconstructions of its equation of state](https://arxiv.org/abs/2503.08658) from combined datasets.
* Advancing our understanding of the early Universe through detailed studies of [21-cm signals from the Cosmic Dawn and Epoch of Reionization](https://arxiv.org/abs/2301.03298), including the development of sophisticated foreground modelling techniques and emulators like [GLOBALEMU](https://arxiv.org/abs/2104.04336).
* Developing new statistical methods for quantifying tensions between cosmological datasets ([Quantifying tensions in cosmological parameters: Interpreting the DES evidence ratio](https://arxiv.org/abs/1902.04029)) and for robust Bayesian model selection ([Bayesian model selection without evidences: application to the dark energy equation-of-state](https://arxiv.org/abs/1506.09024)).
* Exploring fundamental physics questions such as potential [parity violation in the Large-Scale Structure using machine learning](https://arxiv.org/abs/2410.16030).
### Charting the Future: AI-Powered Cosmological Discovery
The Handley Research Group is poised to lead a new era of cosmological analysis, driven by the explosive growth in data from next-generation observatories and transformative advances in artificial intelligence. Our future ambitions are centred on harnessing these capabilities to address the most pressing questions in fundamental physics.
**Strategic Research Pillars:**
* **Next-Generation Simulation-Based Inference (SBI):** We are developing advanced SBI frameworks to move beyond traditional likelihood-based analyses. This involves creating sophisticated codes for simulating [Cosmic Microwave Background (CMB)](https://arxiv.org/abs/1908.00906) and [Baryon Acoustic Oscillation (BAO)](https://arxiv.org/abs/1607.00270) datasets from surveys like DESI and 4MOST, incorporating realistic astrophysical effects and systematic uncertainties. Our AI initiatives in this area focus on developing and implementing cutting-edge SBI algorithms, particularly [neural ratio estimation (NRE) methods](https://arxiv.org/abs/2407.15478), to enable robust and scalable inference from these complex simulations.
* **Probing Fundamental Physics:** Our enhanced analytical toolkit will be deployed to test the standard cosmological model ($\Lambda$CDM) with unprecedented precision and to explore [extensions to Einstein's General Relativity](https://arxiv.org/abs/2006.03581). We aim to constrain a wide range of theoretical models, from modified gravity to the nature of [dark matter](https://arxiv.org/abs/2106.02056) and [dark energy](https://arxiv.org/abs/1701.08165). This includes leveraging data from upcoming [gravitational wave observatories](https://arxiv.org/abs/1803.10210) like LISA, alongside CMB and large-scale structure surveys from facilities such as Euclid and JWST.
* **Synergies with Particle Physics:** We will continue to strengthen the connection between cosmology and particle physics by expanding the [GAMBIT framework](https://arxiv.org/abs/2009.03286) to interface with our new SBI tools. This will facilitate joint analyses of cosmological and particle physics data, providing a holistic approach to understanding the Universe's fundamental constituents.
* **AI-Driven Theoretical Exploration:** We are pioneering the use of AI, including [large language models and symbolic computation](https://arxiv.org/abs/2401.00096), to automate and accelerate the process of theoretical model building and testing. This innovative approach will allow us to explore a broader landscape of physical theories and derive new constraints from diverse astrophysical datasets, such as those from GAIA.
Our overarching goal is to remain at the forefront of scientific discovery by integrating the latest AI advancements into every stage of our research, from theoretical modeling to data analysis and interpretation. We are excited by the prospect of using these powerful new tools to unlock the secrets of the cosmos.
Content generated by [gemini-2.5-pro-preview-05-06](https://deepmind.google/technologies/gemini/) using [this prompt](/prompts/content/index.txt).
Image generated by [imagen-3.0-generate-002](https://deepmind.google/technologies/gemini/) using [this prompt](/prompts/images/index.txt).
```
2. **Paper Metadata:**
```yaml
!!python/object/new:feedparser.util.FeedParserDict
dictitems:
id: http://arxiv.org/abs/2509.24949v1
guidislink: true
link: https://arxiv.org/abs/2509.24949v1
title: Parallel Nested Slice Sampling for Gravitational Wave Parameter Estimation
title_detail: !!python/object/new:feedparser.util.FeedParserDict
dictitems:
type: text/plain
language: null
base: ''
value: Parallel Nested Slice Sampling for Gravitational Wave Parameter Estimation
updated: '2025-09-29T15:45:24Z'
updated_parsed: !!python/object/apply:time.struct_time
- !!python/tuple
- 2025
- 9
- 29
- 15
- 45
- 24
- 0
- 272
- 0
- tm_zone: null
tm_gmtoff: null
links:
- !!python/object/new:feedparser.util.FeedParserDict
dictitems:
href: https://arxiv.org/abs/2509.24949v1
rel: alternate
type: text/html
- !!python/object/new:feedparser.util.FeedParserDict
dictitems:
href: https://arxiv.org/pdf/2509.24949v1
rel: related
type: application/pdf
title: pdf
summary: Inferring parameters and testing hypotheses from gravitational wave signals
is a computationally intensive task central to modern astrophysics. Nested sampling,
a Bayesian inference technique, has become an established standard for this in
the field. However, most common implementations lack the ability to fully utilize
modern hardware acceleration. In this work, we demonstrate that when nested sampling
is reformulated in a natively vectorized form and run on modern GPU hardware,
we can perform inference in a fraction of the time of legacy nested sampling implementations
whilst preserving the accuracy and robustness of the method. This scalable, GPU-accelerated
approach significantly advances nested sampling for future large-scale gravitational-wave
analyses.
summary_detail: !!python/object/new:feedparser.util.FeedParserDict
dictitems:
type: text/plain
language: null
base: ''
value: Inferring parameters and testing hypotheses from gravitational wave signals
is a computationally intensive task central to modern astrophysics. Nested
sampling, a Bayesian inference technique, has become an established standard
for this in the field. However, most common implementations lack the ability
to fully utilize modern hardware acceleration. In this work, we demonstrate
that when nested sampling is reformulated in a natively vectorized form and
run on modern GPU hardware, we can perform inference in a fraction of the
time of legacy nested sampling implementations whilst preserving the accuracy
and robustness of the method. This scalable, GPU-accelerated approach significantly
advances nested sampling for future large-scale gravitational-wave analyses.
tags:
- !!python/object/new:feedparser.util.FeedParserDict
dictitems:
term: astro-ph.IM
scheme: http://arxiv.org/schemas/atom
label: null
- !!python/object/new:feedparser.util.FeedParserDict
dictitems:
term: gr-qc
scheme: http://arxiv.org/schemas/atom
label: null
published: '2025-09-29T15:45:24Z'
published_parsed: !!python/object/apply:time.struct_time
- !!python/tuple
- 2025
- 9
- 29
- 15
- 45
- 24
- 0
- 272
- 0
- tm_zone: null
tm_gmtoff: null
arxiv_comment: To be submitted to SciPost Physics Proceedings (EuCAIFCon 2025)
arxiv_primary_category:
term: astro-ph.IM
authors:
- !!python/object/new:feedparser.util.FeedParserDict
dictitems:
name: David Yallup
- !!python/object/new:feedparser.util.FeedParserDict
dictitems:
name: Metha Prathaban
- !!python/object/new:feedparser.util.FeedParserDict
dictitems:
name: James Alvey
- !!python/object/new:feedparser.util.FeedParserDict
dictitems:
name: Will Handley
author_detail: !!python/object/new:feedparser.util.FeedParserDict
dictitems:
name: Will Handley
author: Will Handley
```
3. **Paper Source (TeX):**
```tex
% =========================================================================
% SciPost LaTeX template
% Version 2024-07
%
% Submissions to SciPost Journals should make use of this template.
%
% INSTRUCTIONS: simply look for the `TODO:' tokens and adapt your file.
% ========================================================================
\documentclass{SciPost}
% Prevent all line breaks in inline equations.
\binoppenalty=10000
\relpenalty=10000
\hypersetup{
colorlinks,
linkcolor={red!50!black},
citecolor={blue!50!black},
urlcolor={blue!80!black}
}
\usepackage[bitstream-charter]{mathdesign}
\urlstyle{same}
% \usepackage{natbib}
\usepackage[capitalize,noabbrev]{cleveref}
\usepackage{booktabs}
\usepackage{subcaption}
% Fix \cal and \mathcal characters look (so it's not the same as \mathscr)
\DeclareSymbolFont{usualmathcal}{OMS}{cmsy}{m}{n}
\DeclareSymbolFontAlphabet{\mathcal}{usualmathcal}
\fancypagestyle{SPstyle}{
\fancyhf{}
\lhead{\colorbox{scipostdeepblue}{\bf \color{white} ~SciPost Physics Proceedings }}
\rhead{{\bf \color{scipostdeepblue} ~Submission }}
\renewcommand{\headrulewidth}{1pt}
\fancyfoot[C]{\textbf{\thepage}}
}
\begin{document}
\pagestyle{SPstyle}
\begin{center}{\Large \textbf{\color{scipostdeepblue}{
%%%%%%%%%% TODO: Write your article's title here
Parallel Nested Slice Sampling for Gravitational Wave Parameter Estimation\\
%%%%%%%%%% END TODO: TITLE
}}}\end{center}
\begin{center}\textbf{
%%%%%%%%%% TODO: AUTHORS
% Write the author list here.
% Use (full) first name (+ middle name initials) + surname format.
% Separate subsequent authors by a comma, omit comma and use "and" for the last author.
% Mark the corresponding author(s) with a superscript symbol in this order
% \star, \dagger, \ddagger, \circ, \S, \P, \parallel, ...
David Yallup\textsuperscript{1$\star$},
Metha Prathaban\textsuperscript{1},
James Alvey\textsuperscript{1} and
Will Handley\textsuperscript{1}
%%%%%%%%%% END TODO: AUTHORS
}\end{center}
\begin{center}
%%%%%%%%%% TODO: AFFILIATIONS
% Write all affiliations here.
% Format: institute, city, country
{\bf 1} Kavli Institute for Cosmology, University of Cambridge, Cambridge, UK
%%%%%%%%%% END TODO: AFFILIATIONS
%%%%%%%%%% TODO: EMAIL
% Provide email address of corresponding author(s)
% \\[\baselineskip]
$\star$ \href{mailto:email1}{\small dy297@cam.ac.uk}\,,\quad
%%%%%%%%%% END TODO: EMAIL
\end{center}
\definecolor{palegray}{gray}{0.95}
\begin{center}
\colorbox{palegray}{
\begin{tabular}{rr}
\begin{minipage}{0.37\textwidth}
\includegraphics[width=60mm]{EuCAIF_logo.png}
\end{minipage}
&
\begin{minipage}{0.5\textwidth}
\vspace{5pt}
\vspace{0.5\baselineskip}
\begin{center} \hspace{5pt}
{\it The 2nd European AI for Fundamental \\Physics Conference (EuCAIFCon2025)} \\
{\it Cagliari, Sardinia, 16-20 June 2025
}
\vspace{0.5\baselineskip}
\vspace{5pt}
\end{center}
\end{minipage}
\end{tabular}
}
\end{center}
\section*{\color{scipostdeepblue}{Abstract}}
\textbf{\boldmath{%
%%%%%%%%%% TODO: ABSTRACT
% Write your abstract here.
Inferring parameters and testing hypotheses from gravitational wave signals is a computationally intensive task central to modern astrophysics. Nested sampling, a Bayesian inference technique, has become an established standard for this in the field. However, most common implementations lack the ability to fully utilize modern hardware acceleration. In this work, we demonstrate that when nested sampling is reformulated in a natively vectorized form and run on modern GPU hardware, we can perform inference in a fraction of the time of legacy nested sampling implementations whilst preserving the accuracy and robustness of the method. This scalable, GPU-accelerated approach significantly advances nested sampling for future large-scale gravitational-wave analyses.
%%%%%%%%%% END TODO: ABSTRACT
}}
\vspace{\baselineskip}
%%%%%%%%%% BLOCK: Copyright information
% This block will be filled during the proof stage, and finilized just before publication.
% It exists here only as a placeholder, and should not be modified by authors.
\noindent\textcolor{white!90!black}{%
\fbox{\parbox{0.975\linewidth}{%
\textcolor{white!40!black}{\begin{tabular}{lr}%
\begin{minipage}{0.6\textwidth}%
{\small Copyright attribution to authors. \newline
This work is a submission to SciPost Phys. Proc. \newline
License information to appear upon publication. \newline
Publication information to appear upon publication.}
\end{minipage} & \begin{minipage}{0.4\textwidth}
{\small Received Date \newline Accepted Date \newline Published Date}%
\end{minipage}
\end{tabular}}
}}
}
%%%%%%%%%% BLOCK: Copyright information
%%%%%%%%%% TODO: LINENO
% For convenience during refereeing we turn on line numbers:
% \linenumbers
% You should run LaTeX twice in order for the line numbers to appear.
%%%%%%%%%% END TODO: LINENO
%%%%%%%%%% TODO: TOC
% Guideline: if your paper is longer that 6 pages, include a TOC
% To remove the TOC, simply cut the following block
% \vspace{10pt}
% \noindent\rule{\textwidth}{1pt}
% \tableofcontents
% \noindent\rule{\textwidth}{1pt}
% \vspace{10pt}
%%%%%%%%%% END TODO: TOC
%%%%%%%%% TODO: CONTENTS
% Write your article contents here, starting from first \section.
% An example structure is given below.
\section{Introduction}
The detection of gravitational waves (GWs) by the LIGO-Virgo-KAGRA collaboration has provided significant advancements in our understanding of the universe, offering new insights into black hole mergers and neutron star coalescences, cosmology, and gravitational theory~\cite{LIGOScientific:2016aoc, LIGO_GWTC1, LIGO_GWTC2, LIGO_GWTC3}. Extracting meaningful information from these signals, however, hinges on robust and efficient inference techniques. Determining the parameters of GW events, such as the masses and spins of the compact objects, and testing competing astrophysical models, often requires computationally intensive Bayesian inference. Nested sampling has emerged as a cornerstone of Bayesian inference in the GW community, providing a powerful framework for both parameter estimation and model comparison. However, despite its robustness and widespread use, nested sampling can be computationally slow, especially when compared to other Markov Chain Monte Carlo (MCMC) methods~\cite{Petrosyan:2022}. This computational bottleneck is a concern, particularly as the volume and complexity of GW data is poised to increase dramatically with next-generation observatories~\cite{Hu:2025}.
To accelerate existing inference tasks and meet the challenges posed by future data, several approaches have been explored. Simulation-Based Inference (SBI) methods, such as neural posterior estimation with implementations like DINGO~\cite{dingo2021, Dax:2022pxd, dingonature}, have demonstrated significant successes in accelerating GW inference and have emerged as a powerful tool in the field. Additionally, efforts have focused on modifying the core nested sampling algorithm, leveraging machine learning tools such as normalizing flows, to accelerate convergence~\cite{Prathaban:2024rmu,Williams:2021qyt}. In this work, we explore a complementary approach: leveraging the parallel processing capabilities of Graphics Processing Units (GPUs) to accelerate nested sampling. While there has been previous work on accelerating MCMC methods on GPUs~\cite{Wong:2022xvh}, we focus on nested sampling. By harnessing the power of modern hardware, we aim to provide an alternative and highly efficient method for GW parameter estimation and model comparison. For current data, this approach can significantly decrease computational demand, enabling the use of a robust and trusted method within the field, but at an accelerated pace.
In this work, we apply a recently developed GPU-accelerated nested sampling framework \cite{yallup2025nested} to the context of GW parameter estimation, complementing the work of Prathaban et al.~\cite{Prathaban:2025qgg}. We focus in this work on demonstrating in a more optimal case, where the likelihood is evaluated on a coarser frequency grid, that we can gain even further computational speedup on real GW parameter estimation problems. We demonstrate that crucially this speedup doesn't just arise from the reduced compute cost of each likelihood call, but the massive parallelism of the core NS algorithm can give dramatic further runtime improvements. This underlines the importance of further developing such accelerated likelihood based inference pipelines for GW inference in the future.
% \cite{cabezas2024blackjax} \cite{yallup2025nested} ~\cite{dingo2021} ~\cite{Dax:2022pxd} \cite{bilby_paper} ~\cite{dynesty} \cite{Prathaban:2024rmu} ~\cite{Bhardwaj:2023xph}
% ~\cite{Polanska:2024zpn}
% ~\cite{LIGOScientific:2016aoc} ~\cite{Khan:2015jqa} ~\cite{Edwards:2023sak} ~\cite{Wong:2022xvh} ~\cite{dingonature}
\section{GPU-Accelerated Nested Sampling}
Nested Sampling has become a prominent method for inference on gravitational wave signals. For example, the \texttt{bilby} software~\cite{bilby_paper} (which itself is a central tool in the field) implements nested sampling as one of its core inference algorithms using the \texttt{dynesty} package~\cite{dynesty}. From the optimization perspective, the utilization of HPC CPU hardware is enhanced through process parallelization as implemented in the parallel \texttt{bilby} extension~\cite{Smith:2019ucc}.
Recently, a reformulation of the nested sampling algorithm has been proposed~\cite{yallup2025nested}, and implemented in the \texttt{blackjax} framework~\cite{cabezas2024blackjax}. We use the recommended combination of algorithm choice and settings identified as \emph{Nested Slice Sampling} (NSS) in \cite{yallup2025nested}. This implementation readily integrates with recent developments in GW modeling and inference that also target GPU hardware, namely fast vectorized waveform generation via the \texttt{ripple} package~\cite{Edwards:2023sak} and likelihood evaluation via the \texttt{jim} software~\cite{wong2023fastgravitationalwaveparameter}. A \emph{bilby-like} kernel has been demonstrated for this task using the same GPU NS framework~\cite{Prathaban:2025qgg}. In this work we deploy the default slice sampling based NS kernel (recommended in Yallup et al.~\cite{yallup2025nested}) as a point of comparison. We also focus particularly on a regime that is complementary to the work of~\cite{Prathaban:2025qgg}, when the likelihood is well parallelised by employing likelihood heterodyning~\cite{Cornish:2021lje}.
% This paper establishes GPU-accelerated nested sampling as a highly competitive update to a classical baseline, offering a compelling alternative to both CPU-based samplers and GPU-accelerated MCMC. Our approach achieves state-of-the-art inference speed and accuracy purely through algorithmic reformulation and hardware acceleration, without reliance on machine learning (ML) enhancements at the parameter sampling level. We do note that one of the distinct advantages over existing nested sampling implementations is that this implementation can uniquely fully exploit ML surrogates at the likelihood level, we consider this an important future direction for the field.
In comparison to \texttt{bilby} (\texttt{dynesty}), the \texttt{blackjax} implementation of Nested Slice Sampling (NSS) is similar at a high level: both implement the classic nested sampling algorithm with an MCMC walk to evolve particles~\cite{dynesty, bilby_paper}. In particular, \texttt{bilby} (\texttt{dynesty}) uses a customized random walk proposal, whereas our sampler uses a slice sampling proposal~\cite{nealslicesampling}. The \texttt{blackjax} implementation, however, executes its slice sampling in a vectorized step across the entire population. Combined with a static memory implementation of the particle update, the entire end-to-end algorithm can then run in GPU memory.
We run with nested sampling hyperparameters, relevant to the \texttt{blackjax} implementation, of: a static population of 3000 live points, with short slice sampling chains of 10$\times$ the number of dimensions in length, and we delete half of the live points at each NS iteration. This represents the default recommended values for the number of particles to delete, the length of the short chains is twice what is usually recommended in \texttt{blackjax}, however it is in-keeping with \texttt{bilby} default values on similar problems. We employ a simple default tuning strategy for the slice sampling chains, using the particle covariance to tune direction proposals. This is troublesome for the wrapped phase and polarization angle parameters in particular, hence the large number of repeats to ensure convergence. Providing better tuning that respects the geometry of the parameter space is an area for future work, but we find that the default tuning is sufficient for the data analysis explored in this work. Being able to delete 1500 live points per iteration highlights the impressive capabilities of a GPU-accelerated nested sampling algorithm, probing parallelism that is largely impossible for CPU implementations.
\section{Application to real data}\label{sec:data}
We validate and benchmark our GPU-accelerated nested sampling
pipeline using real gravitational wave data from the GW150914
event, the first direct detection of gravitational waves from a
binary black hole merger~\cite{LIGOScientific:2016aoc}. This analysis allows us to assess the
performance of our implementation, particularly its runtime and the
effective sample size (ESS). For a direct and fair comparison, we compare to the GPU-accelerated MCMC sampler, FlowMC~\cite{Wong:2022xvh}, which is optimized for the same hardware. We note that it has already been shown that FlowMC (steered via the \texttt{jim} package) agree with the results obtained using \texttt{bilby} in this context~\cite{wong2023fastgravitationalwaveparameter}, and it has been shown that \texttt{blackjax} NS can be brought into nearly exact agreement with \texttt{bilby} when deployed with the same inner kernel~\cite{Prathaban:2025qgg}. We follow mostly the default settings of the \texttt{jim} example script included in the code repository for this event. We increase the number of chains from 500 to 1000, probing similar levels of parallelism to the \texttt{blackjax} implementation, as well as increasing reliable convergence. In both cases we exploit the use of likelihood heterodyning~\cite{Cornish:2021lje}. We fix the same reference parameters used to perform the heterodyning between algorithms, and do not include this in the quoted runtimes. We run both algorithms on a single NVIDIA A100 GPU, with 40GB of memory, and a single CPU core.
We analyze data from the LIGO detectors at Hanford (H1) and
Livingston (L1)~\cite{LIGOScientific:2016aoc}. The IMRPhenomD aligned-spin waveform model~\cite{Khan:2015jqa} is used in this analysis, and we sample
over the resulting binary black hole parameter space. The parameter definitions and the priors used in
the analysis are as listed in~\cite{Prathaban:2025qgg}. We do not include any additional
parameters in the analysis to account for calibration uncertainties, which enables a direct comparison with~\cite{wong2023fastgravitationalwaveparameter, Polanska:2024zpn}.
% \begin{figure}[ht]
% \vskip 0.2in
% \begin{center}
% \centerline{\includegraphics[width=\columnwidth]{figures/GW150914_corner.pdf}}
% \caption{Comparison between the GPU-based \texttt{blackjax} nested sampler and FlowMC for the posterior on the chirp mass, luminosity distance, and sky position in the GW150914 event.}
% \label{reduced_corner}
% \end{center}
% \vskip -0.2in
% \end{figure}
% \begin{table}[ht]
% \caption{Runtime for sampling GW150914, where $*$ indicates values taken from~\cite{wong2023fastgravitationalwaveparameter}}
% \label{sample-table}
% \vskip 0.15in
% \begin{center}
% \begin{small}
% \begin{sc}
% \begin{tabular}{lcccr}
% \toprule
% Algorithm & Runtime (s) & ESS \\
% \midrule
% blackjax nss & 207 & 17490 (7599) \\
% FlowMC & 742 & 13633\\
% bilby$*$ (dynesty) & $1\times10^4$ & 5130\\
% \bottomrule
% \end{tabular}
% \end{sc}
% \end{small}
% \end{center}
% \vskip -0.1in
% \end{table}
\begin{figure}[ht!]
\centering
% Subfigure for the plot
\begin{subfigure}[b]{0.48\linewidth} % Reduced width slightly for safety
\includegraphics[width=\linewidth]{figures/GW150914_corner.pdf}
\caption{Comparison between the GPU-based \texttt{blackjax} nested sampler and FlowMC for the posterior on the chirp mass, luminosity distance, and sky position in the GW150914 event.}
\label{fig:reduced_corner}
% \vspace{-0.1in}
\end{subfigure}% <-- The % is critical to remove the invisible space
\hfill
% Subtable for the table
\begin{subtable}[b]{0.48\linewidth} % Reduced width slightly for safety
\centering
% Use tabular* to force the table to a specific width
\setlength{\tabcolsep}{4pt} % Reduce column padding slightly
\begin{tabular*}{\linewidth}{@{\extracolsep{\fill}} lcr}
\toprule
Algorithm & Runtime (s) & ESS \\
\midrule
\texttt{blackjax} nss & 207 & 17490 (7599) \\ % Abbreviating for space
FlowMC & 742 & 13633\\
bilby$*$ & $10^4$ & 5130\\
\bottomrule
\end{tabular*}
\caption{Runtime for sampling GW150914, where $*$ indicates values taken from~\cite{wong2023fastgravitationalwaveparameter}, the bracket ESS values refer to equal weight samples.}
\label{tab:sample-table}
\vspace{0.5in}
\end{subtable}
\caption{Runtime and posterior inference on GW150914.}
\label{fig:combined_results}
\end{figure}
\begin{figure}
\centering
\includegraphics[width=0.49\linewidth]{figures/scaling_l4_live.pdf}
\hfill
\includegraphics[width=0.49\linewidth]{figures/scaling_l4_n_del.pdf}
\caption{Runtime scaling for nested sampling inference with a heterodyned likelihood on the GW150914 event. Left shows the runtime scaling with a number of deleted particles fixed to half the number of live points, the naive linear scaling expected if the algorithm is not parallelised is shown as a dashed line. Right shows the runtime scaling for a fixed number of 1000 live points as the number of deleted particles is scaled, this time the best case of perfect parallelism is shown as the dashed line.}
\label{fig:scaling}
\end{figure}
We present the runtimes and effective sample sizes (ESS) of the
resulting posterior samples in Table \ref{tab:sample-table}. The
\texttt{blackjax} NSS implementation achieves a runtime of 207
seconds, demonstrating a significant speedup compared to the
CPU-based implementation of \texttt{bilby} (runtime taken
from \cite{wong2023fastgravitationalwaveparameter}), while also
converging almost 3 times as fast as FlowMC. Further, we find that
\texttt{blackjax} NSS achieves a substantially higher ESS per second than both \texttt{bilby} (\texttt{dynesty}) and FlowMC. We evaluate the ESS of FlowMC via the standard measure implemented in the \texttt{arviz} package~\cite{arviz_2019}, and compute the ESS of \texttt{blackjax} nested sampling chain using the \emph{kish} measure as implemented in the \texttt{anesthetic} package~\cite{anesthetic}.
% \footnote{~When compressed to equal weight posterior samples this represents 7599 independent samples}.
This indicates that the \texttt{blackjax} implementation is more
efficient at exploring the posterior distribution per unit of
computational time. Whilst some of the computational cost of FlowMC is amortized in the global density proposal, affording increased efficiency asymptotically, similar schemes have been proposed for nested sampling that could greatly enhance this method in a similar manner~\cite{Prathaban:2024rmu}. The marginalized posteriors are plotted in~\cref{fig:reduced_corner} for a reduced set of the full parameter space that is explored, we note that both algorithms have converged to very similar distributions. Performing some ablations of parameters controlling the runtime suggests that these are conservative, but reliable algorithm hyperparameters for both algorithms on this task.
This demonstrates our GPU-accelerated
nested sampling pipeline as a viable method for robust and efficient
GW parameter estimation, and slice sampling can provide a robust alternative to the standard parallel-walk. We demonstrate the parallel nature of the algorithm in this regime by studying the total runtime on the same parameter estimation problem whilst varying two hyperparameters of the algorithm in \cref{fig:scaling}. We demonstrate that by increasing the size of the live population, or by increasing the deleted fraction of the population, significant gains in runtime are possible. This scaling analysis is run on a single NVIDIA L4 GPU.
Ultimately we chose to focus on parameter estimation as the primary task in this work, but importantly the \texttt{blackjax} nested sampling implementation is itself a classical nested sampling algorithm, and thus can be used to compute the Bayesian evidence for model comparison. Validating the accuracy of this estimation, in light of ML assisted techniques~\cite{Polanska:2024zpn}, alongside exploration of more advanced waveform models and likelihoods, is a highlighted area for future work.
%flowmc hetrodyned sampling time: (11m39+ 43s)
\section{Conclusions}\label{sec:conclusion}
In this work we have demonstrated the application of our GPU-accelerated
nested sampling implementation to the analysis of real gravitational
wave data from the GW150914 event. Our key results, presented in Table
\ref{tab:sample-table} and Figure \ref{fig:reduced_corner}, show a
significant improvement in computational efficiency compared to
established CPU-based methods using \texttt{bilby}, achieving runtime speedups
by two orders of magnitude while maintaining a high Effective Sample Size
(ESS). We draw direct comparison to a similarly GPU-accelerated likelihood based MCMC sampler, FlowMC~\cite{Wong:2022xvh}, and find that the \texttt{blackjax} nested sampling implementation converges in a comparable runtime and yields a higher ESS per second. This is despite limited tuning of the slice sampling kernel which we expect to improve these results even further.
Looking forwards, whilst not explored here, our nested sampling approach also directly yields reliable evidence estimates with informative error bars for no extra computational cost, simplifying the parameter estimation and model
comparison process. These results underscore the potential of our
method to accelerate the analysis of gravitational wave signals,
paving the way for more efficient and comprehensive investigations
of future gravitational wave events.
The impressive parallelism exhibited by GPU nested sampling will be a crucial focus for the broader field of astrophysical inference going forward. As available computational resources shift further towards GPUs, algorithms that can exploit the parallelism opportunities of these devices will be essential. Nested Sampling is already well established as a strong baseline for Bayesian inference across the field, and this work demonstrates that nested sampling is not just a legacy baseline, but a powerful and efficient tool for the future.
\section*{Acknowledgements}
MP is supported by the Harding Distinguished Postgraduate Scholars Programme (HDPSP). JA is supported by a fellowship from the Kavli Foundation. The authors were supported by the research environment and infrastructure of the Handley Lab at the University of Cambridge. We thank the \texttt{jim} and \texttt{ripple} authors for the public codes that were influential to this work.
% TODO: include author contributions
% \paragraph{Author contributions}
% This is optional. If desired, contributions should be succinctly described in a single short paragraph, using author initials.
% TODO: include funding information
% \paragraph{Funding information}
% Authors are required to provide funding information, including relevant agencies and grant numbers with linked author's initials. Correctly-provided data will be linked to funders listed in the \href{https://www.crossref.org/services/funder-registry/}{\sf Fundref registry}.
% \begin{appendix}
% \numberwithin{equation}{section}
% \section{Priors}\label{sec:priors}
% The prior distributions used in the analysis are detailed in
% Table~\ref{tab:priors}. These priors were chosen to enable a direct comparison with~\cite{wong2023fastgravitationalwaveparameter}.
% The luminosity distance prior was taken to be a power law from 1 Mpc to 2000 Mpc, with a power of 2.
% \begin{table}[ht]
% \centering
% \caption{Prior distributions used in the GW150914 analysis.}
% \label{tab:priors}
% \begin{tabular}{lll}
% \toprule
% Parameter & Description & Prior \\
% \midrule
% $M_c$ & Detector-frame chirp mass & Uniform(10, 80) \\
% $q$ & Mass ratio ($m_2/m_1$) & Uniform(0.125, 1.0) \\
% $s_{1z}$ & Aligned spin component 1 & Uniform(-1, 1) \\
% $s_{2z}$ & Aligned spin component 2 & Uniform(-1, 1) \\
% $\iota$ & Inclination angle & Sine(0, $\pi$) \\
% $d_L$ & Luminosity distance & Power Law(1, 2000, 2) \\
% $t_c$ & Coalescence time & Uniform(-0.05, 0.05) \\
% $\phi_c$ & Coalescence phase & Uniform(0, $2\pi$) \\
% $\psi$ & Polarization angle & Uniform(0, $\pi$) \\
% $\alpha$ & Right ascension & Uniform(0, $2\pi$) \\
% $\delta$ & Declination & Cosine($-\pi/2$, $\pi/2$) \\
% \bottomrule
% \end{tabular}
% \end{table}
% \section{Full posterior}
% \begin{figure*}[ht]
% \vskip 0.2in
% \begin{center}
% \centerline{\includegraphics[width=0.825\columnwidth]{figures/GW150914_corner_full.pdf}}
% \caption{Comparison between the GPU-based \texttt{blackjax} nested sampler and FlowMC for the posterior across all parameters describing the GW150914 event.}
% \label{corner}
% \end{center}
% \vskip -0.2in
% \end{figure*}
% \end{appendix}
%%%%%%%%% END TODO: CONTENTS
%%%%%%%%%% TODO: BIBLIOGRAPHY
% Provide your bibliography here. You have two options:
%%% FIRST OPTION
% Write your entries here directly, following the example below, including:
% Author(s), Title, Journal Ref. with year in parentheses at the end, followed by the DOI number.
% \begin{thebibliography}{99}
% \bibitem{1931_Bethe_ZP_71}
% H. A. Bethe, \textit{Zur Theorie der Metalle. i. Eigenwerte und Eigenfunktionen der linearen Atomkette}, Zeit. f{\"u}r Phys. \textbf{71}, 205 (1931), \doi{10.1007\%2FBF01341708}.
% \bibitem{arXiv:1108.2700}
% P. Ginsparg, \textit{It was twenty years ago today...}, (arXiv preprint) \doi{10.48550/arXiv.1108.2700}.
% \end{thebibliography}
%%% SECOND OPTION
% Use your bibtex library, formatted by the SciPost style file.
\bibliography{SciPost_Example_BiBTeX_File.bib}
%%%%%%%%%% END TODO: BIBLIOGRAPHY
\end{document}
```
4. **Bibliographic Information:**
```bbl
\begin{thebibliography}{10}
\providecommand{\url}[1]{\texttt{#1}}
\providecommand{\urlprefix}{URL }
\expandafter\ifx\csname urlstyle\endcsname\relax
\providecommand{\doi}[1]{doi:\discretionary{}{}{}#1}\else
\providecommand{\doi}{doi:\discretionary{}{}{}\begingroup
\urlstyle{rm}\Url}\fi
\providecommand{\eprint}[2][]{\url{#2}}
\bibitem{LIGOScientific:2016aoc}
B.~P. Abbott \emph{et~al.},
\newblock \emph{{Observation of Gravitational Waves from a Binary Black Hole
Merger}},
\newblock Phys. Rev. Lett. \textbf{116}(6), 061102 (2016),
\newblock \doi{10.1103/PhysRevLett.116.061102},
\newblock \eprint{1602.03837}.
\bibitem{LIGO_GWTC1}
B.~Abbott \emph{et~al.},
\newblock \emph{Gwtc-1: A gravitational-wave transient catalog of compact
binary mergers observed by ligo and virgo during the first and second
observing runs},
\newblock Physical Review X \textbf{9}(3) (2019),
\newblock \doi{10.1103/physrevx.9.031040}.
\bibitem{LIGO_GWTC2}
R.~Abbott \emph{et~al.},
\newblock \emph{Gwtc-2: Compact binary coalescences observed by ligo and virgo
during the first half of the third observing run},
\newblock Physical Review X \textbf{11}(2) (2021),
\newblock \doi{10.1103/physrevx.11.021053}.
\bibitem{LIGO_GWTC3}
R.~Abbott \emph{et~al.},
\newblock \emph{Gwtc-3: Compact binary coalescences observed by ligo and virgo
during the second part of the third observing run},
\newblock Physical Review X \textbf{13}(4) (2023),
\newblock \doi{10.1103/physrevx.13.041039}.
\bibitem{Petrosyan:2022}
A.~Petrosyan and W.~J. Handley,
\newblock \emph{Supernest: accelerated nested sampling applied to astrophysics
and cosmology} (2022), \eprint{2212.01760}.
\bibitem{Hu:2025}
Q.~Hu and J.~Veitch,
\newblock \emph{Costs of bayesian parameter estimation in third-generation
gravitational wave detectors: a review of acceleration methods} (2025),
\eprint{2412.02651}.
\bibitem{dingo2021}
M.~Dax, S.~R. Green, J.~Gair, J.~H. Macke, A.~Buonanno and B.~Sch\"olkopf,
\newblock \emph{Real-time gravitational wave science with neural posterior
estimation},
\newblock Phys. Rev. Lett. \textbf{127}, 241103 (2021),
\newblock \doi{10.1103/PhysRevLett.127.241103}.
\bibitem{Dax:2022pxd}
M.~Dax, S.~R. Green, J.~Gair, M.~P\"urrer, J.~Wildberger, J.~H. Macke,
A.~Buonanno and B.~Sch\"olkopf,
\newblock \emph{{Neural Importance Sampling for Rapid and Reliable
Gravitational-Wave Inference}},
\newblock Phys. Rev. Lett. \textbf{130}(17), 171403 (2023),
\newblock \doi{10.1103/PhysRevLett.130.171403},
\newblock \eprint{2210.05686}.
\bibitem{dingonature}
M.~Dax, S.~R. Green, J.~Gair, N.~Gupte, M.~P{\"u}rrer, V.~Raymond,
J.~Wildberger, J.~H. Macke, A.~Buonanno and B.~Sch{\"o}lkopf,
\newblock \emph{Real-time inference for binary neutron star mergers using
machine learning},
\newblock Nature \textbf{639}(8053), 49 (2025).
\bibitem{Prathaban:2024rmu}
M.~Prathaban, H.~Bevins and W.~Handley,
\newblock \emph{{Accelerated nested sampling with $\beta$-flows for
gravitational waves}} (2024),
\newblock \eprint{2411.17663}.
\bibitem{Williams:2021qyt}
M.~J. Williams, J.~Veitch and C.~Messenger,
\newblock \emph{{Nested sampling with normalizing flows for gravitational-wave
inference}},
\newblock Phys. Rev. D \textbf{103}(10), 103006 (2021),
\newblock \doi{10.1103/PhysRevD.103.103006},
\newblock \eprint{2102.11056}.
\bibitem{Wong:2022xvh}
K.~W.~k. Wong, M.~Gabri\'e and D.~Foreman-Mackey,
\newblock \emph{{flowMC: Normalizing flow enhanced sampling package for
probabilistic inference in JAX}},
\newblock J. Open Source Softw. \textbf{8}(83), 5021 (2023),
\newblock \doi{10.21105/joss.05021},
\newblock \eprint{2211.06397}.
\bibitem{yallup2025nested}
D.~Yallup, N.~Kroupa and W.~Handley,
\newblock \emph{Nested slice sampling},
\newblock In \emph{Frontiers in Probabilistic Inference: Learning meets
Sampling} (2025).
\bibitem{Prathaban:2025qgg}
M.~Prathaban, D.~Yallup, J.~Alvey, M.~Yang, W.~Templeton and W.~Handley,
\newblock \emph{{Gravitational-wave inference at GPU speed: A bilby-like nested
sampling kernel within blackjax-ns}} (2025),
\newblock \eprint{2509.04336}.
\bibitem{bilby_paper}
G.~Ashton \emph{et~al.},
\newblock \emph{{BILBY: A user-friendly Bayesian inference library for
gravitational-wave astronomy}},
\newblock Astrophys. J. Suppl. \textbf{241}(2), 27 (2019),
\newblock \doi{10.3847/1538-4365/ab06fc},
\newblock \eprint{1811.02042}.
\bibitem{dynesty}
J.~S. {Speagle},
\newblock \emph{{DYNESTY: a dynamic nested sampling package for estimating
Bayesian posteriors and evidences}},
\newblock MNRAS \textbf{493}(3), 3132 (2020),
\newblock \doi{10.1093/mnras/staa278},
\newblock \eprint{1904.02180}.
\bibitem{Smith:2019ucc}
R.~J.~E. Smith, G.~Ashton, A.~Vajpeyi and C.~Talbot,
\newblock \emph{{Massively parallel Bayesian inference for transient
gravitational-wave astronomy}},
\newblock Mon. Not. Roy. Astron. Soc. \textbf{498}(3), 4492 (2020),
\newblock \doi{10.1093/mnras/staa2483},
\newblock \eprint{1909.11873}.
\bibitem{cabezas2024blackjax}
A.~Cabezas, A.~Corenflos, J.~Lao and R.~Louf,
\newblock \emph{Blackjax: Composable {B}ayesian inference in {JAX}} (2024),
\eprint{2402.10797}.
\bibitem{Edwards:2023sak}
T.~D.~P. Edwards, K.~W.~K. Wong, K.~K.~H. Lam, A.~Coogan, D.~Foreman-Mackey,
M.~Isi and A.~Zimmerman,
\newblock \emph{{Differentiable and hardware-accelerated waveforms for
gravitational wave data analysis}},
\newblock Phys. Rev. D \textbf{110}(6), 064028 (2024),
\newblock \doi{10.1103/PhysRevD.110.064028},
\newblock \eprint{2302.05329}.
\bibitem{wong2023fastgravitationalwaveparameter}
K.~W.~K. Wong, M.~Isi and T.~D.~P. Edwards,
\newblock \emph{Fast gravitational wave parameter estimation without
compromises} (2023), \eprint{2302.05333}.
\bibitem{Cornish:2021lje}
N.~J. Cornish,
\newblock \emph{{Heterodyned likelihood for rapid gravitational wave parameter
inference}},
\newblock Phys. Rev. D \textbf{104}(10), 104054 (2021),
\newblock \doi{10.1103/PhysRevD.104.104054},
\newblock \eprint{2109.02728}.
\bibitem{nealslicesampling}
R.~M. Neal,
\newblock \emph{{Slice sampling}},
\newblock The Annals of Statistics \textbf{31}(3), 705 (2003),
\newblock \doi{10.1214/aos/1056562461}.
\bibitem{Khan:2015jqa}
S.~Khan, S.~Husa, M.~Hannam, F.~Ohme, M.~P\"urrer, X.~Jim\'enez~Forteza and
A.~Boh\'e,
\newblock \emph{{Frequency-domain gravitational waves from nonprecessing
black-hole binaries. II. A phenomenological model for the advanced detector
era}},
\newblock Phys. Rev. D \textbf{93}(4), 044007 (2016),
\newblock \doi{10.1103/PhysRevD.93.044007},
\newblock \eprint{1508.07253}.
\bibitem{Polanska:2024zpn}
A.~Polanska, T.~Wouters, P.~T.~H. Pang, K.~K.~W. Wong and J.~D. McEwen,
\newblock \emph{{Accelerated Bayesian parameter estimation and model selection
for gravitational waves with normalizing flows}},
\newblock In \emph{{38th conference on Neural Information Processing Systems}}
(2024), \eprint{2410.21076}.
\bibitem{arviz_2019}
R.~Kumar, C.~Carroll, A.~Hartikainen and O.~Martin,
\newblock \emph{Arviz a unified library for exploratory analysis of bayesian
models in python},
\newblock Journal of Open Source Software \textbf{4}(33), 1143 (2019),
\newblock \doi{10.21105/joss.01143}.
\bibitem{anesthetic}
W.~Handley,
\newblock \emph{anesthetic: nested sampling visualisation},
\newblock The Journal of Open Source Software \textbf{4}(37), 1414 (2019),
\newblock \doi{10.21105/joss.01414}.
\end{thebibliography}
```
5. **Author Information:**
- Lead Author: {'name': 'David Yallup'}
- Full Authors List:
```yaml
David Yallup:
postdoc:
start: 2021-01-10
thesis: null
original_image: images/originals/david_yallup.jpg
image: /assets/group/images/david_yallup.jpg
links:
ORCiD: https://orcid.org/0000-0003-4716-5817
linkedin: https://www.linkedin.com/in/dyallup/
Metha Prathaban:
phd:
start: 2022-10-01
supervisors:
- Will Handley
thesis: null
partiii:
start: 2020-10-01
end: 2021-06-01
supervisors:
- Will Handley
thesis: Evidence for a Palindromic Universe
original_image: images/originals/metha_prathaban.png
image: /assets/group/images/metha_prathaban.jpg
links:
Harding Scholar: https://www.hardingscholars.fund.cam.ac.uk/metha-prathaban-2022-cohort
GitHub: https://github.com/mrosep
James Alvey:
coi:
start: 2024-10-01
thesis: null
image: https://www.kicc.cam.ac.uk/sites/default/files/styles/inline/public/images/profile/profilepicture-min_1.jpeg?itok=ccmz2RPK
links:
Department webpage: https://www.kicc.cam.ac.uk/staff/dr-james-alvey
Will Handley:
pi:
start: 2020-10-01
thesis: null
postdoc:
start: 2016-10-01
end: 2020-10-01
thesis: null
phd:
start: 2012-10-01
end: 2016-09-30
supervisors:
- Anthony Lasenby
- Mike Hobson
thesis: 'Kinetic initial conditions for inflation: theory, observation and methods'
original_image: images/originals/will_handley.jpeg
image: /assets/group/images/will_handley.jpg
links:
Webpage: https://willhandley.co.uk
```
This YAML file provides a concise snapshot of an academic research group. It lists members by name along with their academic roles—ranging from Part III and summer projects to MPhil, PhD, and postdoctoral positions—with corresponding dates, thesis topics, and supervisor details. Supplementary metadata includes image paths and links to personal or departmental webpages. A dedicated "coi" section profiles senior researchers, highlighting the group’s collaborative mentoring network and career trajectories in cosmology, astrophysics, and Bayesian data analysis.
====================================================================================
Final Output Instructions
====================================================================================
- Combine all data sources to create a seamless, engaging narrative.
- Follow the exact Markdown output format provided at the top.
- Do not include any extra explanation, commentary, or wrapping beyond the specified Markdown.
- Validate that every bibliographic reference with a DOI or arXiv identifier is converted into a Markdown link as per the examples.
- Validate that every Markdown author link corresponds to a link in the author information block.
- Before finalizing, confirm that no LaTeX citation commands or other undesired formatting remain.
- Before finalizing, confirm that the link to the paper itself [2509.24949](https://arxiv.org/abs/2509.24949) is featured in the first sentence.
Generate only the final Markdown output that meets all these requirements.
{% endraw %}