GroNLP
@GroNlp
Natural Language Processing and Computational Linguistics group at the University of Groningen 🐮 [email protected]
[1/]💡New Paper Large reasoning models (LRMs) are strong in English — but how well do they reason in your language? Our latest work uncovers their limitation and a clear trade-off: Controlling Thinking Trace Language Comes at the Cost of Accuracy 📄Link: arxiv.org/abs/2505.22888
✨ New Paper ✨ [1/] Retrieving passages from many languages can boost retrieval augmented generation (RAG) performance, but how good are LLMs at dealing with multilingual contexts in the prompt? 📄 Check it out: arxiv.org/abs/2504.00597 (w/ @AriannaBisazza @raquel_dmg) #NLProc
Last week to apply for our PhD position!
Apply for our PhD position in language acquisition / computational linguistics in Groningen until 24 April! Job ad is here: rug.nl/about-ug/work-…
Modern LLMs "speak" hundreds of languages... but do they really? Multilinguality claims are often based on downstream tasks like QA & MT, while *formal* linguistic competence remains hard to gauge in lots of languages... Meet MultiBLiMP! (joint work w/@JumeletJ & @LAWeissweiler)
✨New paper ✨ Introducing 🌍MultiBLiMP 1.0: A Massively Multilingual Benchmark of Minimal Pairs for Subject-Verb Agreement, covering 101 languages! We present over 125,000 minimal pairs and evaluate 17 LLMs, finding that support is still lacking for many languages. 🧵⬇️
Apply for our PhD position in language acquisition / computational linguistics in Groningen until 24 April! Job ad is here: rug.nl/about-ug/work-…
Interested in a paid PhD position in the Netherlands? Join me, Annemarie van Dooren, and @AriannaBisazza on a project to develop computational models of modal verb acquisition. A job ad will be out soon, but here's a preview: yevgen.web.rug.nl/phd.pdf
📣 Soon-to-open PhD position @GroNlp: Come work with Annemarie van Dooren, @Yevgen_M and myself on a new project bridging Computational Linguistics methods and Language Acquisition questions, with a focus on the learning of modal verbs.
Interested in a paid PhD position in the Netherlands? Join me, Annemarie van Dooren, and @AriannaBisazza on a project to develop computational models of modal verb acquisition. A job ad will be out soon, but here's a preview: yevgen.web.rug.nl/phd.pdf
Fruit of a collaboration started just a few months ago, our investigation of how child-directed Variation Sets help learning in LMs has received a @babyLMchallenge award #CoNLL2024 #EMNLP2024 !! 🎉❤️ W/ amazing co-authors @_akari000 @Rodamille @akiyohukat_u and Yohei Oseki
👶I am happy to announce that our paper "BabyLM Challenge: Exploring the Effect of Variation Sets on Language Model Training Efficiency" received the ✨Outstanding Paper Award✨ at @babyLMchallenge !! #CoNLL2024 #EMNLP2024
The GOLEM Knowledge Graph: Modelling Fiction and Narrative Across Domains GroNLP's Dr. Franziska Pannach is presenting the #GOLEM Ontology and Knowledge Graph at the International Conference Linked Open Data and Literary Studies in Berlin #LODaLS #DH #CLS
🌴We had a great time at #EMNLP2024 presenting our works, meeting old friends, getting to know new people, and winning some prizes (Best Social Impact Award at #EMNLP2024 Main and @babyLMchallenge award #CoNLL2024) 🤩
Being at #EMNL2024 🌴? Come to visit our poster “Language is Scary when Over-Analyzed” (Jasmine 14:00-15:30) We frame misogyny detection as an Argumentative Reasoning task using LLMs. Curious about the results? Come and talk to us @Aramona4
If you're attending #EMNLP2024 don't miss our latest works in interpretability and generation 👇