Machine Learning for Molecules 2026

With the topic “Machine Learning for Molecules", this Summer School 2026 will cover multiple aspects of this young interdisciplinary research area:

• Materials representations and ML­-based materials property prediction
• Atomistic simulations enabled by machine­-learned potentials
• Graph neural networks and inverse molecular design
• Self­-driving labs in chemistry
• LLMs and synthesis prediction

We want to address young researchers at early career stages specifically, i.e. undergraduate students in informatics, chemistry and the material sciences as well as first and second year PhD students. The program will contain both lecture-type presentations as well as interactive formats such as a lab visit, hands-on tutorials, and a poster session in which the participants can present their own research. Along with the scientific program, we will organize a side program consisting of a visit to ZKM (Center for Art and Media Karlsruhe) including a guided tour, a social dinner as well as a public evening including a panel discussion on the topic "Large Language Models in Science - Hype or Future?".

The organizing team (Tobias Schlöder and Pascal Friederich) is happy to assist and answer any questions – don’t hesitate to contact us for more information.

Location and Time

The Summer school will be held in Karlsruhe, Germany from 14 to 18 September 2026.

Program

Confirmed speakers

• Olexandr Isayev (Carnegie Mellon University)
• Rocío Mercado (Chalmers University of Technology)
• ...

Schedule

• Monday 14th: Molecular property prediction
• Tuesday 15th: Self­-driving labs
• Wednesday 16th: Graph neural networks
• Thursday 17th: ML potentials for materials simulations
• Friday 18th: Generative models and LLMs

Panel discussion

As part of the Summer School, we will organise a public panel discussion on Tuesday with the topic "LLMs in Science – hype or future?". Large Language Models are not only becoming a part of our daily lives but also of how we do science. However, it is very hard to estimate to what extent they will influence, accelerate, or even drive scientific progress in the future. We will discuss these and other questions with three panelists, followed by engaging in a conversation with the audience in an informal get-together with snacks and drinks.

Application

Interested candidates are invited to submit an application through this registration link. We will review applications on a continuous basis. If/once an application is approved, further information as well as a link for payment will be provided.

Participants fee

The participation fee covers organization, local support, lunch and drinks during the day, the lab visit, the guided tour at ZKM, and the social dinner. Transport to and from Karlsruhe, as well as accommodation, needs to be individually organized and paid for by the participants.
We will offer reduced prices for undergraduates and PhD students a well as an early bird discount available until June 10th.

Funding

The Summer School Machine Learning for Chemistry is funded by the Carl-Zeiss-Stiftung with additional financial support by the KIT centers MaTeLiS and KCIST.

Previous Summer Schools

This is our fourth consecutive Summer School. You can find a list of all Summer Schools organized by us here.