Linguistic structure · Multilingual NLP · Security and privacy
Johannes Bjerva
Professor of Natural Language Processing · Group Leader
Johannes studies how linguistic structure helps explain the behaviour of multilingual language models, especially their security and privacy risks.
Johannes leads AAU NLP and is Head of the Copenhagen Section of Aalborg University’s Department of Computer Science. His research uses linguistic typology and formal semantics to study model representations, cross-lingual transfer and failure.
He leads projects on multilingual model security and the detection of tampered language models, and co-leads AI:SECURITY with Qiongxiu Li.
Linguistically grounded NLP
Using typology, semantics, and cross-lingual structure to understand how models learn and generalise across languages.
Language-model security
Studying memorisation, inversion, poisoning, and other threats through a multilingual and linguistically informed lens.
Trustworthy language technology
Connecting fundamental NLP to factuality, education, privacy, safety, and reliable deployment in real-world settings.
Selected programmes
Projects and labs
Selected work
5 representative papers
MultiHal: Multilingual Dataset for Knowledge-Graph Grounded Evaluation of LLM Hallucinations
Builds a multilingual multihop benchmark from 25,900 curated knowledge-graph paths for evaluating and reducing hallucinations.
- Knowledge, Evidence & Reasoning
- Multilingual & Lower-Resource NLP
Shared Path: Unraveling Memorization in Multilingual LLMs through Language Similarities
- Security, Privacy & Safety
- Multilingual & Lower-Resource NLP
Large Language Models are Easily Confused: A Quantitative Metric, Security Implications and Typological Analysis
- Security, Privacy & Safety
- Multilingual & Lower-Resource NLP
The Role of Typological Feature Prediction in NLP and Linguistics
- Language, Meaning & Typology
- Multilingual & Lower-Resource NLP
The Meaning Factory: Formal Semantics for Recognizing Textual Entailment and Determining Semantic Similarity
- Language, Meaning & Typology
- Knowledge, Evidence & Reasoning
Supervision and collaboration
Research with Johannes Bjerva
Johannes welcomes projects that connect linguistic analysis with multilingual NLP, model security or privacy.