The CosmoPhysarum project: a statistically calibrated reconstruction of the cosmic web from a bio-inspired algorithm

We present CosmoPhysarum, a GPU-accelerated (CUDA) agent-based code that reconstructs the cosmic web from sparse tracers by emulating the slime mould Physarum polycephalum: agents foraging between galaxies or haloes grow a continuous network that delineates filaments, walls and voids and provides an estimate of the underlying matter field. How faithfully such networks follow the matter field has not been quantified systematically, and CosmoPhysarum is built for the large, reproducible parameter studies this requires. We test it voxel by voxel, in web environment and density, with chance-corrected measures. Grown from fiducial Quijote friends-of-friends haloes with a mean separation of $13.5\,h^{-1}\,{\rm Mpc}$, the unsmoothed reconstruction classifies the web as faithfully as the matter field smoothed over $4$--$6\,h^{-1}\,{\rm Mpc}$, two to three times finer than the tracer separation. At $2.4\,h^{-1}\,{\rm Mpc}$ smoothing, it assigns the correct environment to $75\%$ of the volume and co-locates $81\%$ of the mass, both about $60\%$ of the attainable agreement above chance; it recovers the volumes of voids, walls and filaments within $3\%$ and their masses within $9\%$, and follows the matter density nearly linearly, within $0.1$ dex from $0.3$ to $5.8$ times the mean. Thus the agents turn a strongly biased tracer into a nearly unbiased density estimate with the correct environment proportions. Node cores, however, receive under two thirds of their mass. (Submitted)

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-17
DOI
https://doi.org/10.5281/zenodo.24324773
Primary Topic
Galaxies: Formation, Evolution, Phenomena
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

The CosmoPhysarum project: a statistically calibrated reconstruction of the cosmic web from a bio-inspired algorithm

Jacek Szklarski, Sulona Kandhai, Vincenzo Salzano, Tomasz Krajewski et al.
Zenodo (CERN European Organization for Nuclear Research)
Galaxies: Formation, Evolution, Phenomena
preprint

The CosmoPhysarum project: a statistically calibrated reconstruction of the cosmic web from a bio-inspired algorithm

Jacek Szklarski, Sulona Kandhai, Vincenzo Salzano, Tomasz Krajewski, Tomasz Denkiewicz, Jitendra Joshi
preprint en

Abstract

We present CosmoPhysarum, a GPU-accelerated (CUDA) agent-based code that reconstructs the cosmic web from sparse tracers by emulating the slime mould Physarum polycephalum: agents foraging between galaxies or haloes grow a continuous network that delineates filaments, walls and voids and provides an estimate of the underlying matter field. How faithfully such networks follow the matter field has not been quantified systematically, and CosmoPhysarum is built for the large, reproducible parameter studies this requires. We test it voxel by voxel, in web environment and density, with chance-corrected measures. Grown from fiducial Quijote friends-of-friends haloes with a mean separation of $13.5\,h^{-1}\,{\rm Mpc}$, the unsmoothed reconstruction classifies the web as faithfully as the matter field smoothed over $4$--$6\,h^{-1}\,{\rm Mpc}$, two to three times finer than the tracer separation. At $2.4\,h^{-1}\,{\rm Mpc}$ smoothing, it assigns the correct environment to $75\%$ of the volume and co-locates $81\%$ of the mass, both about $60\%$ of the attainable agreement above chance; it recovers the volumes of voids, walls and filaments within $3\%$ and their masses within $9\%$, and follows the matter density nearly linearly, within $0.1$ dex from $0.3$ to $5.8$ times the mean. Thus the agents turn a strongly biased tracer into a nearly unbiased density estimate with the correct environment proportions. Node cores, however, receive under two thirds of their mass. (Submitted)

Zenodo (CERN European Organization for Nuclear Research)
Galaxies: Formation, Evolution, Phenomena
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon