arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2406.12158cs.CLcs.AI
收藏

大语言模型在因果推断中易犯谬误

LLMs Are Prone to Fallacies in Causal Inference

New York University(纽约大学);University of Michigan(密歇根大学)

Nitish Joshi, Abulhair Saparov, Yixin Wang, He He

AI总结:

本研究针对LLMs因果推断能力是否依赖预训练记忆的问题,通过合成数据微调实验发现其易受提及顺序、时间先后影响产生后此谬误,且难以从反事实关系推断因果,质疑其因果理解能力。

摘要:

近期研究表明,通过提示工程可从大语言模型(LLMs)中有效提取因果事实,助力构建用于因果推断任务的因果图。但目前尚不清楚这一成效是否仅限于模型能记忆的预训练数据中明确提及的因果事实。为此,本研究探讨了一个核心问题:LLMs能否从文本中的其他关系数据推断因果关系?\n为区分记忆的因果事实与推断的因果关系的作用,我们在包含时间、空间和反事实关系的合成数据上微调LLMs,并衡量模型能否据此推断因果关系。研究发现:(a) LLMs易根据文本中两个实体的提及顺序推断因果关系(例如X先于Y被提及就意味着X导致Y);(b) 若随机化提及顺序,LLMs仍会陷入后此谬误,即X在时间上先于Y发生(时间关系)就意味着X导致Y。我们还发现,尽管LLMs能从时间和空间关系中正确推断因果关系的不存在,但它们难以从反事实关系中推断因果关系,这对其因果理解能力提出了质疑。

英文摘要:

Recent work shows that causal facts can be effectively extracted from LLMs through prompting, facilitating the creation of causal graphs for causal inference tasks. However, it is unclear if this success is limited to explicitly-mentioned causal facts in the pretraining data which the model can memorize. Thus, this work investigates: Can LLMs infer causal relations from other relational data in text? To disentangle the role of memorized causal facts vs inferred causal relations, we finetune LLMs on synthetic data containing temporal, spatial and counterfactual relations, and measure whether the LLM can then infer causal relations. We find that: (a) LLMs are susceptible to inferring causal relations from the order of two entity mentions in text (e.g. X mentioned before Y implies X causes Y); (b) if the order is randomized, LLMs still suffer from the post hoc fallacy, i.e. X occurs before Y (temporal relation) implies X causes Y. We also find that while LLMs can correctly deduce the absence of causal relations from temporal and spatial relations, they have difficulty inferring causal relations from counterfactuals, questioning their understanding of causality.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑