The paper introduces an Activation Flow Network that uses L2-norm activation strength at BERT layer 8 to show semantic content words occupy high-activation buckets while structural tokens do not.
Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing (EMNLP) , year=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Towards Explainability of SLMs by investigating Token Level Activation
The paper introduces an Activation Flow Network that uses L2-norm activation strength at BERT layer 8 to show semantic content words occupy high-activation buckets while structural tokens do not.