- SentEval: An Evaluation Toolkit for Universal Sentence Representations . Alexis Conneau and Douwe Kiela . arXiv:1803.05449, 2018.
- Toolkit developed by Facebook AI Research (FAIR) for evaluating the quality of universal sentence representations.
- Useful resources:
- Universal Sentence Encoder . Daniel Cer, Yinfei Yang, Sheng-yi Kong, Nan Hua, Nicole Limtiaco, Rhomni St. John, Noah Constant, Mario Guajardo-Céspedes, Steve Yuan, Chris Tar, Yun-Hsuan Sung, Brian Strope, Ray Kurzweil. arXiv:1803.11175, 2018.
- Sentence encoding developed by Google using multi-task learning.
- Useful resources:
- Learning general purpose distributed sentence representations via large scale multi-task learning . Sandeep Subramanian, Adam Trischler, Yoshua Bengio, Christopher Pal . In ICLR, 2018.
- Another multi-task-learning-based sentence encoding developed by Microsoft research Montreal and Montréal Institute for Learning Algorithms (MILA).
- Useful resources:
- Supervised Learning of Universal Sentence Representations from Natural Language Inference Data . Alexis Conneau, Douwe Kiela, Holger Schwenk, Loic Barrault, Antoine Bordes . In ACL, 2017.
- Supervised learning of sentence representations using Natural Language Inference (NLI) task, developed by Facebook AI Research (FAIR).
- Useful resources:
- Skip-Thought Vectors . Jamie Ryan Kiros, Yukun Zhu, Ruslan Salakhutdinov, Richard S. Zemel, Antonio Torralba, Raquel Urtasun, Sanja Fidler . In NIPS, 2015.
- Unsupervised learning of sentence representations using an encoder-decoder model, where the decoder tries to reconstruct the surrounding sentences of the encoded sentence, like what is done in skip-gram model for word embedding.
- Useful resources: