# "classic" TPS with openpathsampling for LiCl dissociation

```{toctree}
:maxdepth: 1
:caption: LiCl dissociation notebooks

1 pytorch ResNet reduced symmetry functions <1_pytorch_ResNet_reduced_symmetry_functions>
2 emulate production and postprocessing training pytorch pyramid SNN, ResNet top <2_emulate_production_and_postprocessing_training_pytorch_pyramidSNN_ResNet_top>
2 emulate production and postprocessing training pytorch pyramidal ResNet <2_emulate_production_and_postprocessing_training_pytorch_pyramidal_ResNet>

2 committor cross correlation <2_committor_cross_validation>

1 keras ResNet reduced symmetry functions <1_keras_ResNet_reduced_symmetry_functions>
1.5 keras SNN full symmetry functions <1.5_keras_SNN_full_symmetry_functions>
2 emulate production and postprocessing training <2_emulate_production_and_postprocessing_training_keras>
```

This folder contains a number of notebooks showing how to use aimmd to learn solvent-dependent transition probabilities.
Here we use the Behler-Dellago symmetry functions to achieve a permutation-, rotation-, and translation-invariant description of the solvent.