Models and datasets for *Child-directed speech facilitates production, not comprehension, in BabyLMs* (Bunzeck & Zarrieß 2026)
Bastian Bunzeck
bbunzeck
AI & ML interests
Cognitive and usage-based approaches to language modeling, language acquisition in humans and machines, small and efficient language models
Recent Activity
updated a collection 9 days ago
Child-directed speech facilitates production (CoNLL 2026) updated a model 2 months ago
BabyLM-community/deu-baseline-small updated a dataset 6 months ago
bbunzeck/german-babylm-5m-subsetsOrganizations
Word learning in small LMs
Small Language Models Also Work With Small Vocabularies
Models and evaluation data for our 2025 COLING paper (https://aclanthology.org/2025.coling-main.404/).
GPT-wee: How Small Can a Small Language Model Really Get?
Models trained as submission for BabyLM challenge 2023. Paper: https://aclanthology.org/2023.conll-babylm.2
Communicative Baby (BabyLM 2025)
Team Groningen-Bielefeld's submission to BabyLM 2025. (https://aclanthology.org/2025.babylm-main.29/)
German BabyLM
Data that can be used for developing developmentally plausible language models in German.
Fifty shapes of BLiMP: syntactic learning curves in LMs
Models analyzed in our 2024 MILLing paper: https://aclanthology.org/2024.clasp-1.7/
Child-directed speech facilitates production (CoNLL 2026)
Models and datasets for *Child-directed speech facilitates production, not comprehension, in BabyLMs* (Bunzeck & Zarrieß 2026)
Communicative Baby (BabyLM 2025)
Team Groningen-Bielefeld's submission to BabyLM 2025. (https://aclanthology.org/2025.babylm-main.29/)
Word learning in small LMs
German BabyLM
Data that can be used for developing developmentally plausible language models in German.
Small Language Models Also Work With Small Vocabularies
Models and evaluation data for our 2025 COLING paper (https://aclanthology.org/2025.coling-main.404/).
Fifty shapes of BLiMP: syntactic learning curves in LMs
Models analyzed in our 2024 MILLing paper: https://aclanthology.org/2024.clasp-1.7/
GPT-wee: How Small Can a Small Language Model Really Get?
Models trained as submission for BabyLM challenge 2023. Paper: https://aclanthology.org/2023.conll-babylm.2