Ensemble Distillation for Unsupervised Constituency Parsing

Open paper
Benchmark Model Rank Results
constituency-grammar-induction-on-ptbEnsemble (Generative MBR)#1Mean F1 (WSJ): 70.4Max F1 (WSJ): 71.9
constituency-grammar-induction-on-ptbEnsemble (Selective MBR)#5Mean F1 (WSJ): 66.2

Rank counts only results with a code link.