arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2510.15972cs.CLcs.AI
收藏

面向自然语言推理的量子 NLP 模型

Quantum NLP models on Natural Language Inference

Indiana University(印第安纳大学)

Ling Sun, Peter Sullivan, Michael Martin, Yun Zhou

AI总结:

本文在少样本 NLI 中比较 QNLP、混合与经典 transformer 模型,借助 lambeq/DisCoCat 构建量子电路,并提出 IGPP 指标与词簇参数共享架构,证明量子模型参数更少且效率更高。

摘要:

量子自然语言处理(QNLP)通过将组合结构直接嵌入量子电路,为语义建模提供了一种新颖方法。本文研究 QNLP 模型在自然语言推理(NLI)任务中的应用,在受限少样本设置下比较量子模型、混合模型以及基于经典 transformer 的模型。我们使用 lambeq 库和 DisCoCat 框架,为句子对构建参数化量子电路,并针对语义相关性和推理分类对其进行训练。为评估效率,我们提出一种新的信息论指标——单参数信息增益(IGPP),它能够独立于模型规模来量化学习动态。结果表明,量子模型在参数量显著更少的情况下,取得了与经典基线相当的性能。基于量子的模型在推理任务上优于随机初始化的 transformer,并在相关性任务上实现更低的测试误差。此外,量子模型表现出显著更高的逐参数学习效率(最高可达经典对应模型的五个数量级以上),凸显了 QNLP 在低资源、结构敏感场景中的前景。为解决电路级隔离问题并促进参数共享,我们还提出一种新颖的基于聚类的架构,通过将门参数绑定到学习得到的词簇而非独立 token 来提升泛化能力。

英文摘要:

Quantum natural language processing (QNLP) offers a novel approach to semantic modeling by embedding compositional structure directly into quantum circuits. This paper investigates the application of QNLP models to the task of Natural Language Inference (NLI), comparing quantum, hybrid, and classical transformer-based models under a constrained few-shot setting. Using the lambeq library and the DisCoCat framework, we construct parameterized quantum circuits for sentence pairs and train them for both semantic relatedness and inference classification. To assess efficiency, we introduce a novel information-theoretic metric, Information Gain per Parameter (IGPP), which quantifies learning dynamics independent of model size. Our results demonstrate that quantum models achieve performance comparable to classical baselines while operating with dramatically fewer parameters. The Quantum-based models outperform randomly initialized transformers in inference and achieve lower test error on relatedness tasks. Moreover, quantum models exhibit significantly higher per-parameter learning efficiency (up to five orders of magnitude more than classical counterparts), highlighting the promise of QNLP in low-resource, structure-sensitive settings. To address circuit-level isolation and promote parameter sharing, we also propose a novel cluster-based architecture that improves generalization by tying gate parameters to learned word clusters rather than individual tokens.

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。

↑