# Ermakov--Based embeddings as a parameter-free layer for physics-informed neural networks

DevFeed: [Ermakov--Based embeddings as a parameter-free layer for physics-informed neural networks](<https://devfeed.tech/articles/ermakov-based-embeddings-as-a-parameter-free-layer-for-physics-informed-neural-networks-116872.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.03964>)

Author: Roberto Bernal-Jaquez, Roberto Romero Huerta, Alexander Schaum

Published: 2026-11-04T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Neural Network](<https://devfeed.tech/topics/neural-network.md>), [Deep neural networks](<https://devfeed.tech/topics/deep-neural-networks.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>)

Tags: [ablation](<https://devfeed.tech/tags/ablation.md>), [amplitude](<https://devfeed.tech/tags/amplitude.md>), [approximation](<https://devfeed.tech/tags/approximation.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [diffusion](<https://devfeed.tech/tags/diffusion.md>), [embeddings](<https://devfeed.tech/tags/embeddings.md>), [physics-informed](<https://devfeed.tech/tags/physics-informed.md>), [quantum](<https://devfeed.tech/tags/quantum.md>), [scale-function](<https://devfeed.tech/tags/scale-function.md>), [scaling](<https://devfeed.tech/tags/scaling.md>)

## AI overview

The paper encodes the Ermakov-Pinney scale function ρ as a parameter-free input embedding for physics-informed neural networks. An exact moment identity computes ρ in closed form from initial data, without an auxiliary network or knowledge of the solution. The construction covers linear diffusion, diffusion in a time-dependent confining trap, and a quantum oscillator with time-dependent frequency. The useful component differs by family: for linear diffusion, diffusive-time reparameterization alone gives most of the gain; for the trapped diffusion case, no time reparameterization exists, the control degrades accuracy, and an amplitude prefactor is essential (reported 40.8x); for the oscillator, the adiabatic approximation fails and the quadratic phase factor matters (5.7x). The method's extrapolation advantage disappears when the reduced solution does not converge to a fixed point. The authors identify the scaling reduction as classical and the moment identity and comparative analysis as their contribution.

## Source excerpt

arXiv:2611.03964v1 Announce Type: new Abstract: Evolution equations with time-dependent coefficients force a neural solver to spend capacity on a scale that the equation already determines. We show that a single classical object, the Ermakov-Pinney scale function $\rho$, can be written into the input layer of a physics-informed neural network as a parameter-free embedding. One construction covers three problem families through the two signs of the Pinney constant $c$: linear diffusion ($c=-1$, $\omega\equiv0$), diffusion in a time-dependent confining trap ($c=-1$, $\omega\neq0$), and the quantum oscillator with time-dependent frequency ($c=+1$). An exact moment identity gives $\rho$ in closed form from the initial data alone, so no auxiliary network or knowledge of the solution is needed. Our central finding is comparative: each family requires a different component of the construction and has a different honest control. In family I, the diffusive-time reparametrisation alone reproduces most of the gain, so that family is a poor test of the idea and shows why an ablation is mandatory. In family II no time reparametrisation exists, the control degrades accuracy, and the amplitude prefactor becomes essential (40.8x). In family III the adiabatic approximation also fails, and the quadratic phase factor is what matters (5.7x). The method does not help when the reduced solution does not converge to a fixed point: its extrapolation advantage then disappears. The reduction itself is the classical scaling-variable change of the self-similar-asymptotics literature; what is new is the moment identity that makes it architectural for time-dependent coefficients, and the comparative anatomy above.