Postdoc: Explainable AI for Language & Learning

University of Helsinki / General; Finland; On-site; EUR 55,000 - 70,000; Full time

Benefits offered by this job

Occupational health care

Health insurance

Sports facilities

Professional development support

Job summary

University of Helsinki invites applications for postdoctoral researchers in the Language Technology group within the Department of Digital Humanities. The roles focus on Explainable & Trustworthy AI for threat analysis and AI for language teaching, with immediate start and a two‑year initial duration.

You will conduct research, collaborate with interdisciplinary teams, implement algorithms in PyTorch/HuggingFace, and publish at top venues while supporting professional development and relocation

Qualifications

• PhD in Natural Language Processing, ML, CS or related field; must be or near completion.
• Experience with NLP, LLMs, and explainability.
• Strong Python programming skills with PyTorch and Transformers.
• Peer‑reviewed publications required for postdoc appointment.
• Excellent oral and written English.

Responsibilities

• Lead foundational and applied research on explainability and trust in language models.
• Develop evaluation and interpretation methods for threats and risk in LLMs.
• Collaborate with project teams and external partners across disciplines.
• Implement algorithms and data analyses using PyTorch, JAX, transformers.
• Publish and present research at AI/ML/NLP venues and conferences.

Skills

NLP

Machine Learning

Python

PyTorch

HuggingFace Transformers

Academic publications

English proficiency

Education

PhD in NLP/ML/CS

Tools

Python libraries

Job description

University of Helsinki invites applications for postdoctoral researchers in the Language Technology group within the Department of Digital Humanities. The roles focus on Explainable & Trustworthy AI for threat analysis and AI for language teaching, with immediate start and a two‑year initial duration.

You will conduct research, collaborate with interdisciplinary teams, implement algorithms in PyTorch/HuggingFace, and publish at top venues while supporting professional development and relocation