| Website | https://marcel.bollmann.me |
| Scholar | https://www.semanticscholar.org/author/Marcel-Bollmann/34887843 |
| ORCID | https://orcid.org/0000-0003-2598-8150 |
| Pixelfed | https://metapixl.com/mbollmann |

| Website | https://marcel.bollmann.me |
| Scholar | https://www.semanticscholar.org/author/Marcel-Bollmann/34887843 |
| ORCID | https://orcid.org/0000-0003-2598-8150 |
| Pixelfed | https://metapixl.com/mbollmann |
@texttheater This is done by the current create_bibkey method in case of key clashes. If there’s a vella-etal-2024-towards already (I haven’t checked, but it should be the case), the method will add a second word to ensure uniqueness.
I guess paper titles starting with “towards” are just more likely to generate clashes. :)
Godwin's Second Law:
"Anyone who uses 'woke' as a pejorative will turn out to be a fuckhead."
All PhD positions are full-time, salaried positions with pension contributions and benefits such as a yearly wellness allowance (that you can use e.g. for gym memberships or massages).
We're excited to find some new & motivated colleagues. Feel free to ask me about anything you're curious about! 3/3
PhD students will be supervised by Marco Kuhlmann and me. The positions are announced in two different, but closely related projects — Marco will take the lead on the “adaptation and modularisation of LLMs” project, while I will be the lead on the “TrustLLM” project on multilingual LLMs, involving tokenization techniques, addressing grammatical correctness and bias, and others.
Applications within the first project are open now until 2023-10-19, appl. for the other project will open ASAP. 2/
📣 We're hiring up to 3 PhD students for research on large language models!
Specific topics include multilingual LLMs for lower-resourced languages, techniques for tokenization and embedding alignment, modularisation and adaptation of LLMs, and benchmarking and evaluation.
More details (+ how to apply) in this thread and on our website:
https://www.ida.liu.se/divisions/aiics/nlp/phd-student/
I want threads on Mastodon to be called “multi-toots”. So that posters of threads can say of themselves “I contain multi-toots.”
(Which otherwise only works as a complaint to your gastroenterologist.)