The 2025 CLASP Conference on Language models And RePresentations

The 2025 CLASP Conference on Language Models and Representations (CLASP 2025) is an annual research event hosted by the Centre for Linguistic Theory and Studies in Probability at the University of Gothenburg, focusing on the theoretical and empirical underpinnings of modern language models. The conference investigates how neural and statistical models encode, represent, and process linguistic structure.

Topics include probing methods, representation learning, compositional generalization, grounding, interpretability, and the relationship between model representations and human language cognition. The event bridges computational linguistics, cognitive science, and machine learning, making it valuable for researchers in NLP, formal linguistics, and AI who want to deepen understanding of what large language models actually learn. Graduate students and early-career researchers are strongly encouraged to submit and attend.