Propagation and Pitfalls: Reasoning-based Assessment of Knowledge Editing through Counterfactual Tasks

Wenyue Hua, Jiang Guo, Mingwen Dong, Henghui Zhu, Patrick Ng, Zhiguo Wang


Abstract
Current knowledge editing approaches struggle to effectively propagate updates to interconnected facts.In this work, we delve into the barriers that hinder the appropriate propagation of updated knowledge within these models for accurate reasoning. To support our analysis, we introduce a novel reasoning-based benchmark, ReCoE (Reasoning-based Counterfactual Editing dataset), which covers six common reasoning schemes in the real world. We conduct an extensive analysis of existing knowledge editing techniques, including input-augmentation, finetuning, and locate-and-edit methods. We found that all model editing methods exhibit notably low performance on this dataset, especially within certain reasoning schemes. Our analysis of the chain-of-thought responses from edited models indicate that, while the models effectively update individual facts, they struggle to recall these facts in reasoning tasks. Moreover, locate-and-edit methods severely deteriorate the models’ language modeling capabilities, leading to poor perplexity and logical coherence in their outputs.
Anthology ID:
2024.findings-acl.743
Volume:
Findings of the Association for Computational Linguistics ACL 2024
Month:
August
Year:
2024
Address:
Bangkok, Thailand and virtual meeting
Editors:
Lun-Wei Ku, Andre Martins, Vivek Srikumar
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
12503–12525
Language:
URL:
https://aclanthology.org/2024.findings-acl.743
DOI:
Bibkey:
Cite (ACL):
Wenyue Hua, Jiang Guo, Mingwen Dong, Henghui Zhu, Patrick Ng, and Zhiguo Wang. 2024. Propagation and Pitfalls: Reasoning-based Assessment of Knowledge Editing through Counterfactual Tasks. In Findings of the Association for Computational Linguistics ACL 2024, pages 12503–12525, Bangkok, Thailand and virtual meeting. Association for Computational Linguistics.
Cite (Informal):
Propagation and Pitfalls: Reasoning-based Assessment of Knowledge Editing through Counterfactual Tasks (Hua et al., Findings 2024)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-4/2024.findings-acl.743.pdf