This is the collection of the MLatom tutorials for the CZS Summer school 2024 in Karlsruhe. MLatom is a general purpose software platform for AI-enhanced computational chemistry. MLatom has an extensive documentation with many other useful tutorials: https://xacs.xmu.edu.cn/docs/mlatom/.

Pavlo O. Dral's lecture notes related to the tutorials.

• not recommended but if you insist: you can run the notebooks on Google colab.
• strongly recommended to run the tutorials on the XACS cloud where everything is installed: http://XACScloud.com (registration and basic use is free). After the registration, we will give you access to the Jupyter Lab. The Jupyter notebooks will be available in your home folder: jupyter_examples/karlsruhe2024. Click on 'Jupyter Lab' on the left menu bar, then on 'Launch Server', and finally on 'Connect'.
• you can also get support on XACS Slack workspace: https://join.slack.com/t/xacs-support/shared_invite/zt-1gm1lpn68-pReQhfYGu813eCqwmvdGvA (#karlsruhe2024 channel)

Overview of tutorial Jupyter notebooks with links to Google colab versions (feel free to do them in any order):

1. Basic introduction to MLPs MLP_intro.ipynb
2. Testing MLPs test_mlp.ipynb
3. Active learning for MLPs al.ipynb
4. Universal MLP: strengths and pitfalls universal_mlp.ipynb
5. Transfer learning tl.ipynb
6. Delta-learning delta.ipynb
7. Multi-state MLP for excited states and surface-hopping dynamics msani.ipynb

These tutorials are the result of the contributions by:

• Pavlo O. Dral
• Fuchun Ge
• Yifan Hou
• Mikolaj Martyka
• Jakub Martinka