How Pre-trained Language Models Capture Factual Knowledge? A
Causal-Inspired Analysis
- URL: http://arxiv.org/abs/2203.16747v1
- Date: Thu, 31 Mar 2022 02:01:26 GMT
- Title: How Pre-trained Language Models Capture Factual Knowledge? A
Causal-Inspired Analysis
- Authors: Shaobo Li, Xiaoguang Li, Lifeng Shang, Zhenhua Dong, Chengjie Sun,
Bingquan Liu, Zhenzhou Ji, Xin Jiang and Qun Liu
- Abstract summary: We show how PLMs generate missing words by relying on effective clues or shortcut patterns.
We check the words that have three typical associations with the missing words: knowledge-dependent, positionally close, and highly co-occurred.
We conclude that the PLMs capture the factual knowledge ineffectively because of depending on the inadequate associations.
- Score: 43.86843444052966
- License: http://creativecommons.org/licenses/by/4.0/
- Abstract: Recently, there has been a trend to investigate the factual knowledge
captured by Pre-trained Language Models (PLMs). Many works show the PLMs'
ability to fill in the missing factual words in cloze-style prompts such as
"Dante was born in [MASK]." However, it is still a mystery how PLMs generate
the results correctly: relying on effective clues or shortcut patterns? We try
to answer this question by a causal-inspired analysis that quantitatively
measures and evaluates the word-level patterns that PLMs depend on to generate
the missing words. We check the words that have three typical associations with
the missing words: knowledge-dependent, positionally close, and highly
co-occurred. Our analysis shows: (1) PLMs generate the missing factual words
more by the positionally close and highly co-occurred words than the
knowledge-dependent words; (2) the dependence on the knowledge-dependent words
is more effective than the positionally close and highly co-occurred words.
Accordingly, we conclude that the PLMs capture the factual knowledge
ineffectively because of depending on the inadequate associations.
Related papers
- Behavioral Testing: Can Large Language Models Implicitly Resolve Ambiguous Entities? [27.10502683001428]
We analyze current state-of-the-art large language models (LLMs) for their proficiency and consistency in applying their factual knowledge when prompted for entities under ambiguity.
Experiments reveal that LLMs perform poorly with ambiguous prompts, achieving only 80% accuracy.
arXiv Detail & Related papers (2024-07-24T09:48:48Z) - FAC$^2$E: Better Understanding Large Language Model Capabilities by
Dissociating Language and Cognition [57.747888532651]
Large language models (LLMs) are primarily evaluated by overall performance on various text understanding and generation tasks.
We present FAC$2$E, a framework for Fine-grAined and Cognition-grounded LLMs' Capability Evaluation.
arXiv Detail & Related papers (2024-02-29T21:05:37Z) - The Queen of England is not England's Queen: On the Lack of Factual
Coherency in PLMs [2.9443699603751536]
Factual knowledge encoded in Pre-trained Language Models (PLMs) enriches their representations and justifies their use as knowledge bases.
Previous work has focused on probing PLMs for factual knowledge by measuring how often they can correctly predict an object entity given a subject and a relation.
In this work, we consider a complementary aspect, namely the coherency of factual knowledge in PLMs, i.e., how often can PLMs predict the subject entity given its initial prediction of the object entity.
arXiv Detail & Related papers (2024-02-02T14:42:09Z) - How Well Do Large Language Models Understand Syntax? An Evaluation by
Asking Natural Language Questions [25.39259677000101]
This study seeks to explore the question through the lens of syntax.
We craft questions targeting nine syntactic knowledge points that are most closely related to sentence comprehension.
Experiments conducted on 24 large language models (LLMs) suggest that most have a limited grasp of syntactic knowledge.
arXiv Detail & Related papers (2023-11-14T16:30:36Z) - Do Large Language Models Know about Facts? [60.501902866946]
Large language models (LLMs) have recently driven striking performance improvements across a range of natural language processing tasks.
We aim to evaluate the extent and scope of factual knowledge within LLMs by designing the benchmark Pinocchio.
Pinocchio contains 20K diverse factual questions that span different sources, timelines, domains, regions, and languages.
arXiv Detail & Related papers (2023-10-08T14:26:55Z) - Knowledge Rumination for Pre-trained Language Models [77.55888291165462]
We propose a new paradigm dubbed Knowledge Rumination to help the pre-trained language model utilize related latent knowledge without retrieving it from the external corpus.
We apply the proposed knowledge rumination to various language models, including RoBERTa, DeBERTa, and GPT-3.
arXiv Detail & Related papers (2023-05-15T15:47:09Z) - Context-faithful Prompting for Large Language Models [51.194410884263135]
Large language models (LLMs) encode parametric knowledge about world facts.
Their reliance on parametric knowledge may cause them to overlook contextual cues, leading to incorrect predictions in context-sensitive NLP tasks.
We assess and enhance LLMs' contextual faithfulness in two aspects: knowledge conflict and prediction with abstention.
arXiv Detail & Related papers (2023-03-20T17:54:58Z) - Probing Across Time: What Does RoBERTa Know and When? [70.20775905353794]
We show that linguistic knowledge is acquired fast, stably, and robustly across domains. Facts and commonsense are slower and more domain-sensitive.
We believe that probing-across-time analyses can help researchers understand the complex, intermingled learning that these models undergo and guide us toward more efficient approaches that accomplish necessary learning faster.
arXiv Detail & Related papers (2021-04-16T04:26:39Z) - Are Pretrained Language Models Symbolic Reasoners Over Knowledge? [5.480912891689259]
We investigate the two most important mechanisms: reasoning and memorization.
For reasoning, we show that PLMs seem to learn to apply some symbolic reasoning rules correctly but struggle with others, including two-hop reasoning.
For memorization, we identify schema conformity (facts systematically supported by other facts) and frequency as key factors for its success.
arXiv Detail & Related papers (2020-06-18T10:40:37Z)
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.