The NLP group at Linköping University, 
Sweden<https://www.ida.liu.se/divisions/aiics/nlp/>, is currently looking for 
up to 3 PhD students for research on large language models.

Specific research topics include:

* modularisation techniques for multilingual pre-training
* efficient fine-tuning and targeted adaptation to specific tasks
* tokenisation and embedding alignment techniques
* addressing grammatical correctness and bias in pre-training
* benchmarking and evaluation

PhD students will be supervised by Marco 
Kuhlmann<https://liu.se/en/employee/marku61> and Marcel 
Bollmann<https://marcel.bollmann.me/>. All PhD positions are full-term, 
salaried positions with attractive employee benefits and pension contributions.

Linköping University (LiU) is an AI research excellence centre with strong and 
broad research and education in AI from a technical and interdisciplinary 
perspective. LiU is the coordinator of one of the four EU-funded networks of AI 
research excellence centres (TAILOR<https://tailor-network.eu/>) and the host 
of the largest basic research programme in Sweden 
(WASP<https://wasp-sweden.org/>). LiU also hosts the Swedish National Academic 
Infrastructure for Supercomputing in Sweden (NAISS<https://www.naiss.se/>), 
which includes an NVIDIA SuperPOD 
(Berzelius<https://www.nsc.liu.se/systems/berzelius/>).

For detailed information about the positions and instructions on how to apply, 
please see

  https://www.ida.liu.se/divisions/aiics/nlp/phd-student/

Potential candidates are welcome to contact Marco or me for additional 
information.

The application deadline is 2023-10-19.


Kind regards,
Marcel

--
Marcel Bollmann, Dr. phil.
Associate Professor in Natural Language Processing
Department of Computer and Information Science, Linköping University, Sweden
www: https://marcel.bollmann.me/

_______________________________________________
Corpora mailing list -- [email protected]
https://list.elra.info/mailman3/postorius/lists/corpora.list.elra.info/
To unsubscribe send an email to [email protected]

Reply via email to