Dily Duan Yi Ong

Kavli Institute for Cosmology

University of Cambridge

Cambridge, UK

I am a final-year PhD student at the University of Cambridge and a cosmologist specialising in the development of machine-learning-enhanced Bayesian inference tools to understand the structure, evolution, and composition of the universe. Prior to joining Cambridge, I completed my undergraduate and master’s degrees at Imperial College London.

I am the author of unimpeded, a Python package that transforms months of supercomputer calculations into seconds on the laptops of cosmologists and astrophysicists, democratising access to expensive nested sampling chains, enabling cosmological model comparison and observational dataset analysis for researchers worldwide. I am also a contributing author of anesthetic, a Python package for processing cosmological nested sampling and MCMC chains.

Research Interests: Cosmology, Astrophysics, Bayesian Statistics, Machine Learning, Model Comparison, Tension Quantification, Nested Sampling

News

Dec 12, 2025	New paper on arXiv: Signatures of star formation inside galactic outflows - examining local galaxies with powerful AGN to find evidence for star formation within galactic outflows. arXiv:2512.10924

Publications

arXiv
Signatures of star formation inside galactic outflows

Dily Duan Yi Ong, Francesco D’Eugenio, Roberto Maiolino, and 13 more authors

arXiv preprint arXiv:2512.10924, Dec 2025

Abs arXiv Bib HTML PDF

Observations have suggested that galactic outflows contain substantial amounts of dense and clumpy molecular gas, creating favourable conditions for igniting star formation. Indeed, theoretical models and hydrodynamical simulations have suggested that stars could form within galactic outflows, representing a new mode of star-formation that differs significantly from the typical star formation in star forming discs. In this paper, we examine 12 local galaxies with powerful Active Galactic Nuclei and high star-formation rate using spectroscopic data from the X-shooter spectrograph at the Very Large Telescope. We investigate the excitation mechanism and physical properties of these outflows via spatially resolved diagnostic diagrams (along with tests to rule out contribution by shocks and external photoionisation). Out of the seven galaxies with clearly detected outflows, we find robust evidence for star formation within the outflow of one galaxy (IRAS 20551-4250), with two additional...
@article{ong2025starformation, title = {Signatures of star formation inside galactic outflows}, author = {Ong, Dily Duan Yi and D'Eugenio, Francesco and Maiolino, Roberto and Arribas, Santiago and Belfiore, Francesco and Bellocchi, Enrica and Carniani, Stefano and Cazzoli, Sara and Cresci, Giovanni and Fabian, Andrew and Ishibashi, Wako and Mannucci, Filippo and Marconi, Alessandro and Russell, Helen and Sturm, Eckhard and Venturi, Giacomo}, journal = {arXiv preprint arXiv:2512.10924}, year = {2025}, month = dec, }
arXiv
A Bayesian Perspective on Evidence for Evolving Dark Energy

Dily Duan Yi Ong, David Yallup, and Will Handley

arXiv preprint arXiv:2511.10631, Nov 2025

Abs arXiv Bib HTML PDF

The DESI collaboration reports a significant preference for a dynamic dark energy model (w0waCDM) over the cosmological constant (ΛCDM) when their data are combined with other frontier cosmological probes. We present a direct Bayesian model comparison using nested sampling to compute the Bayesian evidence, revealing a contrasting conclusion: for the key combination of the DESI DR2 BAO and the Planck CMB data, we find the Bayesian evidence modestly favours ΛCDM (log-Bayes factor ln B = -0.57±0.26), in contrast to the collaboration’s 3.1σ frequentist significance in favoring w0waCDM.
@article{ong2025bayesian, title = {A Bayesian Perspective on Evidence for Evolving Dark Energy}, author = {Ong, Dily Duan Yi and Yallup, David and Handley, Will}, journal = {arXiv preprint arXiv:2511.10631}, year = {2025}, month = nov, }
arXiv
unimpeded: A Public Grid of Nested Sampling Chains for Cosmological Model Comparison and Tension Analysis

Dily Duan Yi Ong and Will Handley

arXiv preprint arXiv:2511.04661, Nov 2025

Abs arXiv Bib HTML PDF Code

Bayesian inference is central to modern cosmology, yet comprehensive model comparison and tension quantification remain computationally prohibitive for many researchers. To address this, we release unimpeded, a publicly available Python library and data repository providing pre-computed nested sampling and MCMC chains. We apply this resource to conduct a systematic analysis across a grid of eight cosmological models, including ΛCDM and seven extensions, and 39 datasets, including individual probes and their pairwise combinations.
@article{ong2025unimpeded, title = {unimpeded: A Public Grid of Nested Sampling Chains for Cosmological Model Comparison and Tension Analysis}, author = {Ong, Dily Duan Yi and Handley, Will}, journal = {arXiv preprint arXiv:2511.04661}, year = {2025}, month = nov, }
arXiv
unimpeded: A Public Nested Sampling Database for Bayesian Cosmology

Dily Duan Yi Ong and Will Handley

arXiv preprint arXiv:2511.05470, Nov 2025

Abs arXiv Bib HTML PDF Code

Bayesian inference is central to modern cosmology. While parameter estimation is achievable with unnormalised posteriors traditionally obtained via MCMC methods, comprehensive model comparison and tension quantification require Bayesian evidences and normalised posteriors, which remain computationally prohibitive for many researchers. To address this, we present unimpeded, a publicly available Python library and data repository providing DiRAC-funded pre-computed nested sampling and MCMC chains with their normalised posterior samples.
@article{ong2025unimpeded_software, title = {unimpeded: A Public Nested Sampling Database for Bayesian Cosmology}, author = {Ong, Dily Duan Yi and Handley, Will}, journal = {arXiv preprint arXiv:2511.05470}, year = {2025}, month = nov, }