Our collaboration has developed numerous data products of broad interest to the scientific community. Below, we highlight some of our most significant contributions. For a more comprehensive overview, please refer to our publication list:
Manticore
The Manticore Project is constructing the most detailed digital twin of our Universe—akin to a cosmic time machine. Rather than relying on simplified models, we employ state-of-the-art AI and advanced statistical methods to reconstruct the evolution of galaxies, dark matter, and cosmic structures over billions of years. By aligning real astronomical data with unprecedented precision, we are reconstructing the Universe’s history as it truly unfolded.
The project and its data products are accessible on the main webpage: https://cosmictwin.org. Several sub-data products have been derived from the main Markov chain, which represents the initial conditions of the digital twins. Below are those generated through our collaboration. Currently, our data products include:
- Manticore-Local: Initial conditions are inferred within a 1 Gpc box centered on the observer. The analysis is limited to 2M++ data. The catalog of identified clusters is available on VizieR.
- Manticore-Deep: Initial conditions are inferred within a ??? Gpc box. This product incorporates data from the SDSS main, LRG, and BOSS galaxy surveys, as well as 6dFGRS, 2dFGRS, and 2M++.
Cosmic voids
Cosmic voids—vast, underdense regions spanning tens of megaparsecs—present unique challenges for robust statistical identification. Their detection must account for selection effects, observational systematics, structure evolution, and galaxy bias, all of which complicate traditional analyses. The BORG analysis, particularly within the Manticore products, addresses these challenges by reconstructing a self-consistent, observation-free view of the Universe at a fixed epoch. Using this framework, we have conducted cosmic void analyses with multiple void-finding algorithms and generated high-fidelity void catalogs.
Void IDentification and Examination toolkit (VIDE)
VIDE is a widely adopted void finder for identifying cosmic voids in galaxy spectroscopic surveys. Using Manticore-Local data, we have compiled an ensemble of about 100 highly reliable cosmic voids in the Nearby Universe. The full description and data catalogs are available at https://voids.cosmictwin.org. The related paper can be accessed on arXiv: arXiv:2507.06866.
As anti-halos
Another class of void finders involves reversing the initial conditions phase to transform cosmic voids into galaxy clusters within the inverted digital twins. Identifying clusters is more straightforward, as they form spatially compact structures that can be isolated using an isodensity contour in the cosmic web. The particles bound to these structures are then remapped into the non-inverted digital twin simulation to reveal the cosmic void. The cosmic void catalog is available on Zenodo: https://zenodo.org/records/10160612. Each of these voids is extracted from digital twin simulations, which are accessible on the Aquila Cloud instance.
Galaxy clusters
Coming soon…