Worth Their Weights? Evaluating Generative Models for Discourse Parsing

Key points

• The recent availability of high-quality generative language models has made it unclear whether RST parsing is better approached as a generative task.1
• We find that at matched parameter counts, discriminative parsers perform much better than generative ones, though much larger generative models sometimes outperform them.1
• Generative models hold a particular advantage in elementary discourse unit (EDU) segmentation, and we report new state-of-the-art results on the RST-DT benchmark in end-to-end parsing as well as segmentation.1

Sources (1)

• [1]Worth Their Weights? Evaluating Generative Models for Discourse Parsing
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 13, 03:36 AM

  Rhetorical Structure Theory (RST) parsing has historically been dominated by discriminative models based on pretrained encoder-only language models.

  The recent availability of high-quality generative language models has made it unclear whether RST parsing is better approached as a generative task.

Extractive summary: sentences quoted from the sources.