A physics-informed neural network trained on a rarefied gas flow produces weights that look Gaussian and random, showing no obvious internal imprint of the Boltzmann equation it was trained to reproduce.
This makes them erratic and hard to predict, but also capable of emergent phenomena—complex behavior arising from simple rules, biology being a goldmine of such instances
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
physics.comp-ph 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Is the end of Insight in Sight ?
A physics-informed neural network trained on a rarefied gas flow produces weights that look Gaussian and random, showing no obvious internal imprint of the Boltzmann equation it was trained to reproduce.