We present DiscoSense, a benchmark for commonsense reasoning via understanding a wide variety of discourse connectives. We generate compelling distractors in DiscoSense using Conditional Adversarial Filtering, an extension of Adversarial Filtering that employs conditional generation. We show that state-of-the-art pre-trained language models struggle to perform well on DiscoSense, which makes this dataset ideal for evaluating next-generation commonsense reasoning systems.
DiscoSense: Commonsense Reasoning with Discourse Connectives
DiscoSense, a benchmark for evaluating commonsense reasoning through discourse connectives, uses Conditional Adversarial Filtering to generate challenging distractors, revealing limitations in state-of-the-art pre-trained language models.
- Year
- 2022
- Venue
- arXiv 2022
- Authors
- 2
- Hosting
- Abstract onlyARXIV-DEFAULT
Cite
Notes
Only stored in your browser.
Attribution
- Abstract & full text
- arxiv.org/abs/2210.12478ARXIV-DEFAULT
- TL;DR
- Semantic Scholar