"Assistant Professor in NLP"

AarhusNLP is seeking a fixed-term (three-year) Assistant Professor in NLP. The position is set to begin 1 October or as soon as possible thereafter. We are a research group based at the Center for Humanities Computing, Aarhus University, focusing on investigating and developing NLP and AI technologies. Our research areas include model evaluation and post-training, representation learning, and the development of open-source software libraries for NLP. We are dedicated to delivering state-of-the-art language technology tailored specifically for the Scandinavian context. The assistant professorship is funded by the Danish Ministry of Digital Affairs’ AI Initiative Danish Foundation Models project, the Lex.llm project with Lex – Denmark’s National Encyclopedia, and the Danish National Research Foundation’s TEXT Center.

Position Summary

The successful candidate will work on alignment, preference optimization, and evaluation of LLMs for mid- to low-resource languages. The main goal is to develop advanced post-training techniques that align pre-trained transformer-based LLMs with cultural values and linguistic norms. By focusing on low-resource languages, the research should address challenges such as data scarcity, cultural sensitivity, inclusivity, and the need for robust preference optimization methods that go beyond standard fine-tuning.

Key research objectives include:

• Developing efficient algorithms: exploring and designing training strategies (e.g., supervised finetuning, reinforcement learning, or new alignment protocols) or inference-time scaling methods suitable for low-resource settings, such as Nordic languages.
• Preference optimization and alignment: embedding usecase-specific values and preferences into LLMs to ensure respectful and context-aware language generation. These may include cultural norms, ethical considerations, and societal values.
• Evaluation and quality assurance: accurately measuring model performance on nondifferentiable properties, including user satisfaction, cultural appropriateness, and domain-specific quality measures (e.g., information retrieval or question-answering for Nordic applications).

Teaching:

• Teaching the course Natural Language Processing at the Cognitive Science program.
• Complete the University Pedagogical Programme as part of employment.

Additional Responsibilities

Apart from research and teaching activities, the successful candidate will have 20% of their time dedicated to administrative tasks:

• Manage daily administrative duties within AarhusNLP, supporting the group’s operations and research efforts.
• Proactively identify, pursue, and contribute to securing external funding to sustain and grow AarhusNLP’s research initiatives.

Qualifications:

• A Ph.D. or equivalent in Computational Linguistics, Computer Science, or a related field (e.g., Cognitive Science), with specialization in NLP, Machine Learning, or a similar area.
• Experience and interest in leading largescale open collaborations toward shared project goals.
• Expertise in LLM architectures and training paradigms (transformer models, finetuning strategies, RLHF, etc.).
• Experience with multilingual and lowresource NLP.
• Programming skills in Python and proficiency in deep learning frameworks (e.g., PyTorch, TensorFlow) and their use in HPC environments.
• A strong or emerging publication record in NLP, ML, or related disciplines.
• Excellent written and verbal communication skills in English. Knowledge of Danish or another Nordic language is a plus, but not required.
• Engagement in opensource communities is helpful, though not mandatory.
• Proven experience in evaluating language models and related systems.

AarhusNLP is a research group at Aarhus University located at Center for Humanities Computing. We investigate and develop NLP and AI tools focusing on model evaluation and post-training, representation learning, and development of software libraries for NLP researchers. We are particularly committed to developing state-of-the-art language technology for the Scandinavian countries.