Abstract
We present a visualisation tool which aims to illuminate the inner workings of an LSTM model for question answering. It plots heatmaps of neurons’ firings and allows a user to check the dependency between neurons and manual features. The system possesses an interactive web-interface and can be adapted to other models and domains.- Anthology ID:
- D18-2006
- Volume:
- Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing: System Demonstrations
- Month:
- November
- Year:
- 2018
- Address:
- Brussels, Belgium
- Editors:
- Eduardo Blanco, Wei Lu
- Venue:
- EMNLP
- SIG:
- SIGDAT
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 30–35
- Language:
- URL:
- https://aclanthology.org/D18-2006
- DOI:
- 10.18653/v1/D18-2006
- Cite (ACL):
- Ekaterina Loginova and Günter Neumann. 2018. An Interactive Web-Interface for Visualizing the Inner Workings of the Question Answering LSTM. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, pages 30–35, Brussels, Belgium. Association for Computational Linguistics.
- Cite (Informal):
- An Interactive Web-Interface for Visualizing the Inner Workings of the Question Answering LSTM (Loginova & Neumann, EMNLP 2018)
- PDF:
- https://preview.aclanthology.org/nschneid-patch-3/D18-2006.pdf