This repository includes the code related to the "On the Curious Case of l2 norm of Sense Embeddings" paper.

Please install the requirements using:

pip install -r requirements.txt

To train the static sense embeddings using GloVe and Word2Vec, you may use:

python train-glove-or-word2vec-for-sense.py

To train static word embeddings using BERT, you may use:

python embed_annotations-words.py

To train static sense embeddings using LMMS on SemCor only, you may use:

python embed_annotations-sense.py

To reproduce the results of MFS prediction task, you may run:

python mfs.py