TNG50 stellar disks are grouped into four j-types from sAM surface density maps, revealing a redshift-dependent sequence from irregular to barred forms driven by gas content and V/σ.
Title resolution pending
44 Pith papers cite this work, alongside 816 external citations. Polarity classification is still indexing.
44
Pith papers citing it
816
external citations · Crossref
citation-role summary
dataset 1
citation-polarity summary
roles
dataset 1polarities
use dataset 1representative citing papers
Four parameters suffice to describe dust attenuation curve diversity in TNG simulations, yielding a new symbolic-regression model that recovers curves and fluxes better than existing parameterizations while linking parameters to SFR surface density, metallicity, and geometry.
RUBIES JWST sample shows U increases with redshift and sSFR, decreases with mass, rising by a factor of ~4 from z=2 to z=6 at fixed mass and sSFR, with 0.3 dex systematic uncertainty from photoionization model range.
Lumen modeling of IllustrisTNG50 shows that high ionization parameters from massive star clusters plus enhanced nitrogen abundances are needed to reproduce the extreme [OIII]/Hβ, [OIII]/[OII], and [NII]/Hα ratios seen in z>3 galaxies.
A local Tremaine-Weinberg framework integrates the continuity equation over flexible loops to measure galactic pattern speeds, recovering standard methods as special cases and validated on IllustrisTNG simulations.
IllustrisTNG yields γ=2.23±0.20 for luminosity density evolution that explains the Tolman and distance-duality test signals in standard cosmology.
A homogenized ICL definition applied to Horizon-AGN, TNG100, Gizmo-Simba and Hydrangea yields consistent z=0 fractions of 0.1-0.2 with no significant redshift evolution and dominant contributions from satellites of 10^10.5-10^11.5 solar masses.
Information-theoretic analysis of TNG50 simulations finds high mutual information (0.4-0.8) between bar and spiral parameters and comparable transfer entropy (0.33-0.42) in both directions, indicating mutual co-regulation.
JWST measurements of pitch angles in 593 spiral galaxies to z=3.5 show no overall redshift evolution but reveal correlations with mass and sSFR only below z=1.25, implying a transition from locally driven to globally regulated spiral arms.
Mutual information analysis of TNG50 simulations shows gravitational potential and total energy retain merger mass and infall time information longest, while radial velocity loses it within ~5 Gyr, with washout depending on radius, merger age, and mass.
In TNG-50, 80% of Milky Way-mass galaxies align their present-day angular momentum with the orbital angular momentum of their most massive merger, and 81% of their stellar halos rotate prograde relative to the disk.
TNG50 MW analogues reproduce global soft X-ray luminosity, inner surface brightness, emission measure and O VII absorption but show too-steep radial decline in X-ray brightness and 65% lower O VIII absorption than observed, indicating overly central feedback.
SHAMe-SF modeling of small-scale DESI ELG clustering delivers 6% precision on σ8 and Ωm h², matching full DR1 results with 1% volume.
Machine learning on cosmological simulations achieves 91-94% accuracy classifying over-massive versus under-massive SMBH growth regimes from LSST photometry, with 83-89% cross-simulation transfer accuracy driven primarily by host galaxy colors.
Milky Way-mass dark matter density profiles in IllustrisTNG are largely insensitive to astrophysics and cosmology variations, dominated by halo-to-halo variance instead.
New JWST data on 23 galaxies at 0.5<z<1.7 show median dark matter fraction of 0.63 at effective radius with 0.2 dex scatter, and a mix of rising, flat, and falling rotation curves.
TNG50 simulation analysis finds early-forming halos show higher stellar mass, metallicity, lower CGM gas mass and sSFR at z=0, with mass-dependent CGM metallicity reversal and cold-gas kinematics differences.
The TNG SAM reproduces TNG hydro simulation gas and metal flows plus galaxy and halo properties within 30% accuracy out to z=6 via five targeted updates to the Santa Cruz SAM calibrated on stellar feedback-dominated galaxies.
Barred galaxies transition from actively star-forming at z~1-2 to quiescent at z~0, with the fraction of quiescent galaxies hosting bars increasing steeply over cosmic time.
PRFM-vol and PRFM-int subgrid models, calibrated from TIGRESS, produce shorter gas depletion times than IllustrisTNG and match TIGRESS star formation rates at high resolution while remaining robust at lower resolutions.
In TNG50, compact dwarf satellites (log M_star 8.4-9.2) form via DM-rich gas inflows in low-merger environments, tidal stripping for DM-poor cases, and ram-pressure starbursts for some metal-rich ones.
TNG-Cluster simulations find that in galaxy cluster centers turbulence accounts for under half the total velocity dispersion (typically 50-75 km/s), is mostly subsonic, provides sub-percent pressure support, and is primarily driven by SMBH feedback.
Resolved photometry of four high-redshift quiescent galaxies reveals negative color gradients that lower estimated stellar masses by 0.1 dex relative to slit measurements, reducing model tensions under an age-driven interpretation.
Simulations of evolving galaxies show strong L_nu-SFR and L_nu-V_rot correlations up to z~3, with turbulent magnetic fields dominating at low redshift and large-scale fields growing in importance at higher redshift.
citing papers explorer
-
Interpreting "Interpretability" and Explaining "Explainability" in Machine Learning in Physics
The paper defines interpretability as model structural transparency and explainability as scientific content mapping, discusses their trade-offs, and frames both as deliberate modeling choices for ML in physics.