Following the publication of our CoNLL 2026 paper that introduces a method for learning large-scale computational construction grammars from semantically annotated corpora, we release a series of grammars and accompanying grammar networks accessible through an interactive iPython notebook: https://gitlab.ai.vub.ac.be/ehai/pyfcg/-/blob/main/docs/source/walkthrough_tutorials/fcg-propbank.ipynb

The accompanying paper has been published online in the ACL Anthology now: https://aclanthology.org/2026.conll-main.13/

Please reach out if you have any further questions or you’d like to use the constructional embeddings in your own research!