Showing Your Work Doesn’t Always Work

Raphael Tang, Jaejun Lee, Ji Xin, Xinyu Liu, Yaoliang Yu, Jimmy Lin


Abstract
In natural language processing, a recently popular line of work explores how to best report the experimental results of neural networks. One exemplar publication, titled “Show Your Work: Improved Reporting of Experimental Results” (Dodge et al., 2019), advocates for reporting the expected validation effectiveness of the best-tuned model, with respect to the computational budget. In the present work, we critically examine this paper. As far as statistical generalizability is concerned, we find unspoken pitfalls and caveats with this approach. We analytically show that their estimator is biased and uses error-prone assumptions. We find that the estimator favors negative errors and yields poor bootstrapped confidence intervals. We derive an unbiased alternative and bolster our claims with empirical evidence from statistical simulation. Our codebase is at https://github.com/castorini/meanmax.
Anthology ID:
2020.acl-main.246
Volume:
Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics
Month:
July
Year:
2020
Address:
Online
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
2766–2772
Language:
URL:
https://aclanthology.org/2020.acl-main.246
DOI:
10.18653/v1/2020.acl-main.246
Bibkey:
Cite (ACL):
Raphael Tang, Jaejun Lee, Ji Xin, Xinyu Liu, Yaoliang Yu, and Jimmy Lin. 2020. Showing Your Work Doesn’t Always Work. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 2766–2772, Online. Association for Computational Linguistics.
Cite (Informal):
Showing Your Work Doesn’t Always Work (Tang et al., ACL 2020)
Copy Citation:
PDF:
https://preview.aclanthology.org/ingestion-script-update/2020.acl-main.246.pdf
Video:
 http://slideslive.com/38929111
Code
 castorini/meanmax