We are excited to announce an opening for a Research Assistant position focused on Language Models (LLM) and Knowledge Graphs, commencing in November. Join our dedicated team and be at the forefront of cutting-edge research in this rapidly evolving domain.

This research project is associated with IntelliLung project which aims to reduce lung injuries by providing automated suggestions to clinicians for mechanically ventilated patients in the ICU. Biased datasets can degrade the performance of the trained algorithms. Considering the importance of safety for this application, it is critical to identify these biases, discover their source and develop strategies to mitigate them. Also the aim is to discover biases that can help us structure and reduce complexity of the algorithms. This project is supported by interdisciplinary teams and aims to identify biases not only from the Machine Learning (ML) / Reinforcement Learning (RL) perspective but also using the domain knowledge provided by clinicians.

In case of any questions or interest, please contact by sending your CV and study transscript:

Mohammad Hamza:  yousuf@infai.org

Sahar Vahdati: sahar.vahdati@tu-dresden.de

The performance of a large language model (LLM) is sensitive to the way it is prompted. Automated prompt engineering methods aim to find suitable prompts for a given task by sampling several prompts and evaluating them. Existing automatic prompt engineering methods do not generate sufficiently diverse sample prompts or rely on several meta-prompting tricks to achieve the desired results. In this thesis, we will use a method for prompt selection to directly optimise diversity and estimated performance by exploiting so called determinental point processes. The thesis will involve comparisons of this technique to state-of-the-art prompt engineering methods such as PromptBreeder from DeepMind.

Requirements: