Rethinking Self-Attention: Towards Interpretability in Neural Parsing

Open paper
Benchmark Model Rank Results
constituency-parsing-on-penn-treebankLabel Attention Layer + HPSG + XLNet#3F1 score: 96.38
dependency-parsing-on-penn-treebankLabel Attention Layer + HPSG + XLNet#1LAS: 96.26UAS: 97.42POS: 97.3

Rank counts only results with a code link.