Rectified sigmoid (hard sigmoid) activation is reported to cut PINN solution errors by about an order of magnitude on two ODE benchmarks, but the result may be an interpolation artifact because the paper never discloses the source of the initialization data.
A Modified Physics Informed Neural Networks for Solving the Partial Differential Equation with Conservation Laws,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
math.NA 1years
2024 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
About rectified sigmoid function for enhancing the accuracy of Physics-Informed Neural Networks
Rectified sigmoid (hard sigmoid) activation is reported to cut PINN solution errors by about an order of magnitude on two ODE benchmarks, but the result may be an interpolation artifact because the paper never discloses the source of the initialization data.