UniArk: Improving Generalisation and Consistency for Factual Knowledge Extraction through Debiasing
Preprint
Abstract
Several recent papers have investigated the potential of language models as knowledge bases as well as the existence of severe biases when extracting factual knowledge. In this work, we focus on the factual probing performance over unseen prompts from tuning, and using a probabilistic view we show the inherent misalignment between pre-training and downstream tuning objectives in language models for probing knowledge. We hypothesize that simultaneously debiasing these objectives can be the key to generalisation over unseen prompts. We propose an adapter-based framework, UniArk, for generalised and consistent factual knowledge extraction through simple methods without introducing extra parameters. Extensive experiments show that UniArk can significantly improve the model's out-of-domain generalisation as well as consistency under various prompts. Additionally, we construct ParaTrex, a large-scale and diverse dataset for measuring the inconsistency and out-of-domain generation of models. Further, ParaTrex offers a reference method for constructing paraphrased datasets using large language models.Keywords
All Related Versions
This publication has 18 references indexed in Scilit:
- A divide and conquer framework for Knowledge EditingKnowledge-Based Systems, 2023
- Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language ProcessingACM Computing Surveys, 2023
- Measuring and Improving Consistency in Pretrained Language ModelsTransactions of the Association for Computational Linguistics, 2021
- Sustainable Modular Debiasing of Language Models2021
- Knowledgeable or Educated Guess? Revisiting Language Models as Knowledge Bases2021
- Prefix-Tuning: Optimizing Continuous Prompts for Generation2021
- AutoPrompt: Eliciting Knowledge from Language Models with Automatically Generated Prompts2020
- Exploiting Linked Data and Knowledge Graphs in Large Organisations2017
- Reasoning Web: Logical Foundation of Knowledge Graph Construction and Query Answering2017
- WikidataCommunications of the ACM, 2014