Untying the Reversal Curse via Bidirectional Language Model Editing

• URL: http://arxiv.org/abs/2310.10322v2
• Date: Sat, 12 Oct 2024 03:31:13 GMT
• Title: Untying the Reversal Curse via Bidirectional Language Model Editing
• Authors: Jun-Yu Ma, Jia-Chen Gu, Zhen-Hua Ling, Quan Liu, Cong Liu,
• Abstract summary: Large language models (LLMs) store massive factual knowledge within their parameters. LLMs are prone to hallucinate unintended text due to false or outdated knowledge. We study bidirectional language model editing to assess if edited LLMs can recall the editing knowledge bidirectionally.
• Score: 41.040662400025184
• License:
• Abstract: Recent studies have demonstrated that large language models (LLMs) store massive factual knowledge within their parameters. But existing LLMs are prone to hallucinate unintended text due to false or outdated knowledge. Since retraining LLMs is resource intensive, there has been a growing interest in the concept of model editing. Despite the emergence of benchmarks and approaches, these unidirectional editing and evaluation have failed to explore the reversal curse. Intuitively, if "The capital of France is" is edited to be a counterfact "London" within a model, then it should be able to naturally reason and recall the reverse fact, i.e., "London is the capital of" followed by "France" instead of "England". In this paper, we study bidirectional language model editing, aiming to provide rigorous model editing evaluation to assess if edited LLMs can recall the editing knowledge bidirectionally. A new evaluation metric of reversibility is introduced, and a benchmark dubbed as Bidirectional Assessment for Knowledge Editing (BAKE) is constructed to evaluate the reversibility of edited models in recalling knowledge in the reverse direction of editing. We surprisingly observe that while current editing methods and LLMs can effectively recall editing facts in the direction of editing, they suffer serious deficiencies when evaluated in the reverse direction. To mitigate the reversal curse, a method named Bidirectionally Inversible Relationship moDeling (BIRD) is proposed. A set of editing objectives that incorporate bidirectional relationships between subject and object into the updated model weights are designed. Experiments show that BIRD improves the performance of four representative LLMs of different sizes via question answering and judgement.

Related papers

• Fundamental Problems With Model Editing: How Should Rational Belief Revision Work in LLMs? [61.68363765350178]
  This paper critiques the standard formulation of the model editing problem and proposes a formal testbed for model editing research. We first describe 12 open problems with model editing, based on challenges with (1) defining the problem, (2) developing benchmarks, and (3) assuming LLMs have editable beliefs in the first place. Next, we introduce a semi-synthetic dataset for model editing based on Wikidata, where we can evaluate edits against labels given by an idealized Bayesian agent.
  arXiv  Detail & Related papers  (2024-06-27T17:33:03Z)
• Editing Conceptual Knowledge for Large Language Models [65.38231526537476]
  This paper pioneers the investigation of editing conceptual knowledge for Large Language Models (LLMs) We construct a novel benchmark dataset ConceptEdit and establish a suite of new metrics for evaluation. experimental results reveal that, although existing editing methods can efficiently modify concept-level definition to some extent, they also have the potential to distort the related instantial knowledge.
  arXiv  Detail & Related papers  (2024-03-10T16:57:10Z)
• On the Robustness of Editing Large Language Models [57.477943944826904]
  Large language models (LLMs) have played a pivotal role in building communicative AI, yet they encounter the challenge of efficient updates. This work seeks to understand the strengths and limitations of editing methods, facilitating practical applications of communicative AI.
  arXiv  Detail & Related papers  (2024-02-08T17:06:45Z)
• Emptying the Ocean with a Spoon: Should We Edit Models? [8.545919917068273]
  We call into question the recently popularized method of direct model editing as a means of correcting factual errors in LLM generations. We contrast model editing with three similar but distinct approaches that pursue better defined objectives.
  arXiv  Detail & Related papers  (2023-10-18T13:38:03Z)
• Unveiling the Pitfalls of Knowledge Editing for Large Language Models [41.83423510576848]
  It is still unclear whether knowledge editing might introduce side effects that pose potential risks or not. This paper pioneers the investigation into the potential pitfalls associated with knowledge editing for Large Language Models. Experimental results vividly demonstrate that knowledge editing might inadvertently cast a shadow of unintended consequences.
  arXiv  Detail & Related papers  (2023-10-03T15:10:46Z)
• Cross-Lingual Knowledge Editing in Large Language Models [73.12622532088564]
  Knowledge editing has been shown to adapt large language models to new knowledge without retraining from scratch. It is still unknown the effect of source language editing on a different target language. We first collect a large-scale cross-lingual synthetic dataset by translating ZsRE from English to Chinese.
  arXiv  Detail & Related papers  (2023-09-16T11:07:52Z)
• Eva-KELLM: A New Benchmark for Evaluating Knowledge Editing of LLMs [54.22416829200613]
  Eva-KELLM is a new benchmark for evaluating knowledge editing of large language models. Experimental results indicate that the current methods for knowledge editing using raw documents are not effective in yielding satisfactory results.
  arXiv  Detail & Related papers  (2023-08-19T09:17:19Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.