About The Position

As Sweden's national center for applied AI, AI Sweden is looking for a master thesis student to join their team. This project focuses on automated evaluation for tracking LLM post-training progress, specifically addressing the biases in "LLM-as-a-judge" paradigms, particularly for non-English languages like Swedish. The thesis will explore whether providing an LLM judge with few-shot, human-calibrated Swedish examples can effectively decouple fluency from adequacy, aligning automated Elo scores with human expert ratings. The goal is to refine and validate the JudgeArena framework for Swedish post-training evaluation.

Requirements

  • Ongoing Master’s studies in Computer Science, Data Science, Machine Learning, Engineering Physics, or a related quantitative field.
  • Proficiency in Python.
  • Hands-on experience with modern deep learning frameworks (PyTorch, Hugging Face ecosystem).
  • Familiarity with LLM post-training alignment (e.g., SFT, DPO, RLHF/RLVR) or context-extension.
  • Comfort running distributed GPU training in Linux/HPC environments.

Nice To Haves

  • Curiosity and self-driven attitude.
  • Eagerness to work at the frontier of open-weight European AI research (LLMs).
  • Thrive on empirical discovery, design rigorous experiments, and let data challenge assumptions.

Responsibilities

  • Review automated evaluation paradigms, LLM-as-a-judge biases, and metrics distinguishing fluency from adequacy.
  • Annotate a high-quality Swedish dataset with separate scores for fluency and adequacy.
  • Use annotated data as calibration prompts for the Judge model in JudgeArena.
  • Correlate the calibrated JudgeArena outputs against human judgments on a holdout set of Prelude 9B's generated responses.
  • Quantify the improvement in Pearson/Spearman correlation.

Benefits

  • Work alongside leading AI scientists and change leaders.
  • Opportunity to drive research questions with long shelf-life and wide applicability to Swedish industry and the public sector.
  • Aim for publications at competitive venues.
  • Celebrate a culture of research excellence.
  • Experience an organization positioned at the sweet spot of governmental influence and startup agility.
  • Be in close contact with government, academia, and private and public sectors.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service