Do you want to learn more about AI for chemical sciences?

The Digital Molecular Design and Fabrication Institute (DigiFAB) and the FoNS Data Science Champions invite third and fourth-year undergraduate students, postgraduate taught and research students, postdocs and early career researchers interested in the area of large language models to participate in this free hackathon!


The Hackathon Challenge

Can large language models be used to design chemical reactions and predict materials' properties?

Teams will compete to use large language models (like ChatGPT) to solve these global challenges. Existing knowledge in chemistry is not necessary, and the Datathon Challenge is a great opportunity to get to know others interested in this area by working together! You can apply either as a team of five people or as an individual and be assigned a team by the Hackathon organisers.

For teams already familiar with AI, the morning session will involve fine-tuning a large language model on chemical datasets. In the afternoon, all teams are invited to use the models for solving chemistry tasks (no previous AI expertise is required).

Prizes:

  • First place team: £100 per person
  • Runner up team: £75 per person
  • Third place team: £50 per person

 

When and where is the hackathon?

20 July, 9.30–17:30
Blackett 320, South Kensington campus. 

Schedule: 

9.30  - 10.00      Welcome and introduction 
10.00 - 12.30     Morning session - Retrosynthesis planning using LLMs
12.30 - 13.00     Lunch - open to all participants
13.00 - 14.00     Invited speaker - Dr Kevin Jablonka and Dr Michael Pieler – LLMs for Molecular and Materials Science

14.00 - 14.30     Break

14.30 - 16.30     Afternoon session - Prompt engineering & knowledge extraction
16.30 - 17.00     Closing address and awards


Registration closed on 30 June.

If you have any question, please email Ester Buchaca-Domingo.

DS