Abstract
Input optimization methods, such as Google Deep Dream, create interpretable representations of neurons for computer vision DNNs. We propose and evaluate ways of transferring this technology to NLP. Our results suggest that gradient ascent with a gumbel softmax layer produces n-gram representations that outperform naive corpus search in terms of target neuron activation. The representations highlight differences in syntax awareness between the language and visual models of the Imaginet architecture.- Anthology ID:
- W18-5437
- Volume:
- Proceedings of the 2018 EMNLP Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP
- Month:
- November
- Year:
- 2018
- Address:
- Brussels, Belgium
- Editors:
- Tal Linzen, Grzegorz Chrupała, Afra Alishahi
- Venue:
- EMNLP
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 325–327
- Language:
- URL:
- https://aclanthology.org/W18-5437
- DOI:
- 10.18653/v1/W18-5437
- Cite (ACL):
- Nina Poerner, Benjamin Roth, and Hinrich Schütze. 2018. Interpretable Textual Neuron Representations for NLP. In Proceedings of the 2018 EMNLP Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP, pages 325–327, Brussels, Belgium. Association for Computational Linguistics.
- Cite (Informal):
- Interpretable Textual Neuron Representations for NLP (Poerner et al., EMNLP 2018)
- PDF:
- https://preview.aclanthology.org/naacl24-info/W18-5437.pdf
- Code
- NPoe/input-optimization-nlp + additional community code