We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CL

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computation and Language

Title: Doing Good or Doing Right? Exploring the Weakness of Commonsense Causal Reasoning Models

Abstract: Pretrained language models (PLM) achieve surprising performance on the Choice of Plausible Alternatives (COPA) task. However, whether PLMs have truly acquired the ability of causal reasoning remains a question. In this paper, we investigate the problem of semantic similarity bias and reveal the vulnerability of current COPA models by certain attacks. Previous solutions that tackle the superficial cues of unbalanced token distribution still encounter the same problem of semantic bias, even more seriously due to the utilization of more training data. We mitigate this problem by simply adding a regularization loss and experimental results show that this solution not only improves the model's generalization ability, but also assists the models to perform more robustly on a challenging dataset, BCOPA-CE, which has unbiased token distribution and is more difficult for models to distinguish cause and effect.
Comments: ACL2021, Main Conference, Short Paper
Subjects: Computation and Language (cs.CL)
Cite as: arXiv:2107.01791 [cs.CL]
  (or arXiv:2107.01791v1 [cs.CL] for this version)

Submission history

From: Mingyue Han [view email]
[v1] Mon, 5 Jul 2021 05:08:30 GMT (1274kb)

Link back to: arXiv, form interface, contact.