How Does Data Corruption Affect Natural Language Understanding Models? A Study on GLUE datasets
Abstract
A central question in natural language understanding (NLU) research is whether high performance demonstrates the models' strong reasoning capabilities. We present an extensive series of controlled experiments where pre-trained language models are exposed to data that have undergone specific corruption transformations. These involve removing instances of specific word classes and often lead to non-sensical sentences. Our results show that performance remains high on most GLUE tasks when the models are fine-tuned or tested on corrupted data, suggesting that they leverage other cues for prediction even in non-sensical contexts. Our proposed data transformations can be used to assess the extent to which a specific dataset constitutes a proper testbed for evaluating models' language understanding capabilities.
- Publication:
-
arXiv e-prints
- Pub Date:
- January 2022
- DOI:
- 10.48550/arXiv.2201.04467
- arXiv:
- arXiv:2201.04467
- Bibcode:
- 2022arXiv220104467T
- Keywords:
-
- Computer Science - Computation and Language
- E-Print:
- *SEM 2022 camera ready version