RNN language models trained using the architecture of Gulordava et al (2018), on a Wikipedia corpus augmented with sentences containing filler-gap dependencies (clefting and topicalization). These models can be loaded using Pytorch using the same instructions as the Gulordava model. For more information about our implementation, and to generate the datasets these models were trained on, see our repo.

If you are using these models or our implementation, please cite: Katherine Howitt, Sathvik Nair, Allison Dods, & Robert Melvin Hopkins (2024). Generalizations across filler-gap dependencies in neural language models Conference on Natural Language Learning.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference API
Unable to determine this model's library. Check the docs .