Abstract
We present a human-in-the-loop dashboard tailored to diagnosing potential spurious features that NLI models rely on for predictions. The dashboard enables users to generate diverse and challenging examples by drawing inspiration from GPT-3 suggestions. Additionally, users can receive feedback from a trained NLI model on how challenging the newly created example is and make refinements based on the feedback. Through our investigation, we discover several categories of spurious correlations that impact the reasoning of NLI models, which we group into three categories: Semantic Relevance, Logical Fallacies, and Bias. Based on our findings, we identify and describe various research opportunities, including diversifying training data and assessing NLI models’ robustness by creating adversarial test suites.- Anthology ID:
- 2023.acl-demo.44
- Volume:
- Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)
- Month:
- July
- Year:
- 2023
- Address:
- Toronto, Canada
- Editors:
- Danushka Bollegala, Ruihong Huang, Alan Ritter
- Venue:
- ACL
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 463–470
- Language:
- URL:
- https://aclanthology.org/2023.acl-demo.44
- DOI:
- 10.18653/v1/2023.acl-demo.44
- Cite (ACL):
- Robin Chan, Afra Amini, and Mennatallah El-Assady. 2023. Which Spurious Correlations Impact Reasoning in NLI Models? A Visual Interactive Diagnosis through Data-Constrained Counterfactuals. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations), pages 463–470, Toronto, Canada. Association for Computational Linguistics.
- Cite (Informal):
- Which Spurious Correlations Impact Reasoning in NLI Models? A Visual Interactive Diagnosis through Data-Constrained Counterfactuals (Chan et al., ACL 2023)
- PDF:
- https://preview.aclanthology.org/improve-issue-templates/2023.acl-demo.44.pdf