The Paul Scherrer Institute PSI is the largest research institute for natural and engineering sciences within Switzerland. We perform cutting-edge research in the fields of future technologies, energy and climate, health innovation and fundamentals of nature. By performing fundamental and applied research, we work on sustainable solutions for major challenges facing society, science and economy. PSI is committed to the training of future generations. Therefore, about one quarter of our staff are post-docs, post-graduates or apprentices. Altogether, PSI employs 2200 people. 

For the Large Research Facilities Division we are looking for a we are looking for a

Trainee for the indexing of scientific documents with the aid of large language models (LLMs)

Your tasks

We are using retrieval augmented generation (RAG) to support a large language model (LLM) answering questions related to our accelerators. Source documents include scientific publications, as well as internal documents. Many of these documents include images, tables, and equations, which are not directly accessible to be indexed in a language-based embedding.

You will help address these issues by:

  • Support the interpretation of scientific publications and internal notes by large language models
  • Assess possibilities to index images, tables, and equations
  • Store embeddings in a vector data base
  • Use a large language model to search this data base, and answer user questions
  • Build a user interface for this model
The vector data base and the large language model run locally on hardware located at PSI.

Your profile

  • You are a student in computer science or similar, and you are at least in your penultimate year of undergraduate study (minimum 4 semesters)
  • You have experience running a large language model locally, or you use an API to access a cloud-based LLM
  • You are an open-minded person who enjoys working in an international team
  • You have not yet completed your Master’s thesis

We offer

Our institution is based on an interdisciplinary, innovative and dynamic collaboration. 

The contract will be limited to 3 months.

For further information, please contact Dr Rasmus Ischebeck, phone +41 56 310 55 35 or email rasmus.ischebeck@psi.ch 

Please submit your application online for the position as a Trainee (index no. 8123-T1).

Paul Scherrer Institut, Human Resources Management, Ladina Goerre, 5232 Villigen PSI, Switzerland