Improving Syntactic Probing Correctness and Robustness with Control Tasks

Syntactic probing methods have been used to examine whether and how pre-trained language models (PLMs) encode syntactic relations. However, the probing methods are usually biased by the PLMs' memorization of common word co-occurrences, even if they do not form syntactic relations. This paper presents a random-word-substitution and random-labelmatching control task to reduce these biases and improve the robustness of syntactic probing methods. Our control tasks are also shown to notably improve the consistency of probing results between different probing methods and make the methods more robust with respect to the text attributes of the probing instances. Our control tasks make syntactic probing methods better at reconstructing syntactic relations and more generalizable to unseen text domains. Our experiments show that our proposed control tasks are effective on different PLMs, probing methods, and syntactic relations.