Comparing Test Sets with Item Response Theory

06/01/2021 ∙ by Clara Vania, et al. ∙ 7

Recent years have seen numerous NLP datasets introduced to evaluate the performance of fine-tuned models on natural language understanding tasks. Recent results from large pretrained models, though, show that many of these datasets are largely saturated and unlikely to be able to detect further progress. What kind of datasets are still effective at discriminating among strong models, and what kind of datasets should we expect to be able to detect future improvements? To measure this uniformly across datasets, we draw on Item Response Theory and evaluate 29 datasets using predictions from 18 pretrained Transformer models on individual test examples. We find that Quoref, HellaSwag, and MC-TACO are best suited for distinguishing among state-of-the-art models, while SNLI, MNLI, and CommitmentBank seem to be saturated for current strong models. We also observe span selection task format, which is used for QA datasets like QAMR or SQuAD2.0, is effective in differentiating between strong and weak models.

READ FULL TEXT
POST COMMENT

Comments

There are no comments yet.

Authors

page 5

Code Repositories

nlu-test-sets

Analysis of NLU test sets with IRT


view repo
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.