# A Descriptor Is All You Need: Accurate Machine Learning of Nonadiabatic Coupling Vectors

arXiv: https://arxiv.org/abs/2505.23344 · HTML: https://mlsanity.com/paper/2505.23344/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 7 of its 123 references that are in the map.
- First posted on arXiv 29 May 2025.

## Builds on (most cited first, 7 of 7)

- [SchNetPack 2.0: A neural network toolbox for atomistic machine learning](https://mlsanity.com/paper/2212.05517/) (2022)
- [Combining SchNet and SHARC: The SchNarc machine learning approach for excited-state dynamics](https://mlsanity.com/paper/2002.07264/) (2020)
- [MLatom 3: Platform for machine learning-enhanced computational chemistry simulations and workflows](https://mlsanity.com/paper/2310.20155/) (2023)
- [Machine learning enables long time scale molecular photodynamics simulations](https://mlsanity.com/paper/1811.09112/) (2018)
- [Excited state, non-adiabatic dynamics of large photoswitchable molecules using a chemically transferable machine learning potential](https://mlsanity.com/paper/2108.04879/) (2021)
- [Neural networks and kernel ridge regression for excited states dynamics of CH2NH2^+: From single-state to multi-state representations and multi-property machine learning models](https://mlsanity.com/paper/1912.08484/) (2019)
- [Machine Learning Nonadiabatic Dynamics: Eliminating Phase Freedom of Nonadiabatic Couplings with the State-Intraction State-Averaged Spin-Restricted Ensemble-Referenced Kohn-Sham Approach](https://mlsanity.com/paper/2410.22801/) (2024)

## Abstract (arXiv)

Nonadiabatic couplings (NACs) play a crucial role in modeling photochemical and photophysical processes with methods such as the widely used fewest-switches surface hopping (FSSH). There is therefore a strong incentive to machine learn NACs for accelerating simulations. However, this is challenging due to NACs' vectorial, double-valued character and the singularity near a conical intersection seam. For the first time, we design NAC-specific descriptors based on our domain expertise and show that they allow learning NACs with never-before-reported accuracy of R2 exceeding 0.99. The key to success is also our new ML phase-correction procedure. We demonstrate the efficiency and robustness of our approach on a prototypical example of fully ML-driven FSSH simulations of fulvene targeting the SA-2-CASSCF(6,6) electronic structure level. This ML-FSSH dynamics leads to an accurate description of S1 decay while reducing error bars by allowing the execution of a large ensemble of trajectories. Our implementations are available in open-source MLatom.