Spring kickoff and new student projects
OdenseNLP has started the spring semester with a new set of MSc student projects focused on Danish NLP, evaluation pipelines, and model transparency. We are excited to collaborate wi...
OdenseNLP
Safe, Efficient and Open Natural Language Processing @ University of Southern Denmark
Latest news
See all posts
NordicBench preprint is now available
We released the first preprint of NordicBench, a benchmark suite designed for low-resource Scandinavian NLP tasks. The benchmark includes baselines, training recipes, and evaluation ...
OdenseNLP co-organizes clinical NLP workshop
Members of OdenseNLP co-organized a regional workshop on clinical NLP in collaboration with partners from Danish hospitals and universities. The workshop highlighted practical challe...
Current focuses
Low-resource NLP
Language technologies for low-resource langauges, particularly Danish and neighboring Scandinavian languages.
Efficient NLP
Fast and efficient NLP architectures and methods.
AI Safety & Interpretability
Making AI systems more safe, trustworthy, and interpretable.
Repositories summary
See all repositoriesDaLA
Source code for creating the Danish Corpus of Linguistic Acceptability, designed to evaluate Danish linguistic acceptability with real-world errors.
brainsurgery
Swiss-army-knife style toolkit for scripted tensor surgery on model checkpoints, with YAML plans, CLI workflows, and a Web UI.
DeToNATION
A communication framework that optimizes distributed AI training by reducing latency bottlenecks and handling heterogeneous clusters.
Datasets summary
See all datasetsDaLA
Danish linguistic acceptability dataset with corrupted and non-corrupted sentences, published as ~8.68k examples with train/val/test and full-train...
SDU-Daisy
Danish-culture benchmark dataset based on the Danish Culture Canon, with 746 closed question-answer pairs for evaluating LLM cultural understanding.