Neha Srikanth


2022

pdf
Partial-input baselines show that NLI models can ignore context, but they don’t.
Neha Srikanth | Rachel Rudinger
Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies

When strong partial-input baselines reveal artifacts in crowdsourced NLI datasets, the performance of full-input models trained on such datasets is often dismissed as reliance on spurious correlations. We investigate whether state-of-the-art NLI models are capable of overriding default inferences made by a partial-input baseline. We introduce an evaluation set of 600 examples consisting of perturbed premises to examine a RoBERTa model’s sensitivity to edited contexts. Our results indicate that NLI models are still capable of learning to condition on context—a necessary component of inferential reasoning—despite being trained on artifact-ridden datasets.

2021

pdf
Elaborative Simplification: Content Addition and Explanation Generation in Text Simplification
Neha Srikanth | Junyi Jessy Li
Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021