collaborators

8 papers

cs.CL2026

Child-directed speech facilitates production, not comprehension, in BabyLMs

Bastian Bunzeck, Sina Zarrieß

Recent studies suggest that child-directed speech is not conducive to language learning in BabyLMs. However, current evaluations focus predominantly on comprehension and not produc…

cs.CL2026

The Frequency Confound in Language-Model Surprisal and Metaphor Novelty

Omar Momen, Sina Zarrieß

Language-model (LM) surprisal is widely used as a proxy for contextual predictability and has been reported to correlate with metaphor novelty judgments. However, surprisal is tigh…

cs.CL2026

How Hypocritical Is Your LLM judge? Listener-Speaker Asymmetries in the Pragmatic Competence of Large Language Models

Judith Sieker, Sina Zarrieß

Large language models (LLMs) are increasingly studied as repositories of linguistic knowledge. In this line of work, models are commonly evaluated both as generators of language an…

cs.CL2025

Dialogue Is Not Enough to Make a Communicative BabyLM (But Neither Is Developmentally Inspired Reinforcement Learning)

Francesca Padovani, Bastian Bunzeck, Manar Ali +4

We investigate whether pre-training exclusively on dialogue data results in formally and functionally apt small language models. Based on this pre-trained llamalogue model, we empl…

cs.CL2025

Do Construction Distributions Shape Formal Language Learning In German BabyLMs?

Bastian Bunzeck, Daniel Duran, Sina Zarrieß

We analyze the influence of utterance-level construction distributions in German child-directed/child-available speech on the resulting word-level, syntactic and semantic competenc…

cs.CL2025

Subword models struggle with word learning, but surprisal hides it

Bastian Bunzeck, Sina Zarrieß

We study word learning in subword and character language models with the psycholinguistic lexical decision task. While subword LMs struggle to discern words and non-words with high…