Amnesic Probing: Behavioral Explanation with Amnesic Counterfactuals

Yanai Elazar, Shauli Ravfogel, Alon Jacovi, Yoav Goldberg


Abstract
A growing body of work makes use of probing in order to investigate the working of neural models, often considered black boxes. Recently, an ongoing debate emerged surrounding the limitations of the probing paradigm. In this work, we point out the inability to infer behavioral conclusions from probing results, and offer an alternative method that focuses on how the information is being used, rather than on what information is encoded. Our method, Amnesic Probing, follows the intuition that the utility of a property for a given task can be assessed by measuring the influence of a causal intervention that removes it from the representation. Equipped with this new analysis tool, we can ask questions that were not possible before, for example, is part-of-speech information important for word prediction? We perform a series of analyses on BERT to answer these types of questions. Our findings demonstrate that conventional probing performance is not correlated to task importance, and we call for increased scrutiny of claims that draw behavioral or causal conclusions from probing results.1
Anthology ID:
2021.tacl-1.10
Volume:
Transactions of the Association for Computational Linguistics, Volume 9
Month:
Year:
2021
Address:
Cambridge, MA
Editors:
Brian Roark, Ani Nenkova
Venue:
TACL
SIG:
Publisher:
MIT Press
Note:
Pages:
160–175
Language:
URL:
https://aclanthology.org/2021.tacl-1.10
DOI:
10.1162/tacl_a_00359
Bibkey:
Cite (ACL):
Yanai Elazar, Shauli Ravfogel, Alon Jacovi, and Yoav Goldberg. 2021. Amnesic Probing: Behavioral Explanation with Amnesic Counterfactuals. Transactions of the Association for Computational Linguistics, 9:160–175.
Cite (Informal):
Amnesic Probing: Behavioral Explanation with Amnesic Counterfactuals (Elazar et al., TACL 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.tacl-1.10.pdf
Video:
 https://aclanthology.org/2021.tacl-1.10.mp4