A geometric physics-informed machine learning inference for the neutron star maximum mass and the inverse problem

Rounak Mukherjee, Ritam Mallick · 2026

Read A geometric physics-informed machine learning inference for the neutron star maximum mass and the inverse problem by Rounak Mukherjee, Ritam Mallick (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-18191.txt.

Description

Rounak Mukherjee and one co-author ask where neutron stars end and black holes begin. Their 2026 paper notes the boundary is open, a gap born of ignorance about matter squeezed past nuclear density inside neutron star cores. The team attacks the question with a physics-informed learning model guided by sky observations.

Trained on a broad, neutral set of matter equations, the Transformer model reads two shape traits of the mass-radius curve, called front and back bending. It first predicts peak mass and radius from those bends, then works backward to the sound-speed profile inside. The model says heavy neutron stars arise when sound speed peaks at low density, with strong back bending and an early switch to quark matter. It puts the peak mass near two and a half Suns, with smallest radius near eleven and a half kilometers at one point four solar masses.

Catalog record

CREATORRounak Mukherjee, Ritam Mallick
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE39 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602-18191

Browse the full library · About the Tower of Babel