基于函数梯度引导的语言模型学习概率逻辑程序
Learning Probabilistic Logic Programs with Functional Gradient Guided Language Models
- Technische Universität Darmstadt（达姆施塔特工业大学）
- The University of Texas at Dallas（德克萨斯大学达拉斯分校）
- AT&T CDO（AT&T首席数字官办公室）
- Oregon state University（俄勒冈州立大学）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
提出神经符号框架grasp，将关系结构学习转化为函数梯度提升，用LLM处理难的内部搜索，在Tox21等四个基准上优于多种基线，生成可解释带权规则集成。
AI中文摘要：
声明式逻辑程序通过将依赖关系表示为带权组合规则，为编码关系结构与神经符号推理提供了强大且可解释的抽象。然而，从数据中归纳这类程序本质上十分困难，瓶颈在于符号搜索空间的组合爆炸。大型语言模型（LLM）近来成为强大的假设生成器，但单独使用时，它们缺乏可靠合成适配复杂关系分布的有效程序所需的系统归纳推理能力。我们提出grasp（概率逻辑程序的梯度增强合成，Gradient-boosted Synthesis of Probabilistic logic programs），这是一个神经符号框架，将关系结构学习转化为函数梯度提升过程，其中弱学习器为一阶规则，难以处理的内部搜索则委托给LLM提议预言机。我们在四个关系基准上评估grasp，涵盖分子毒性预测（Tox21）、诱变及引文匹配（Cora），结果显示它优于纯符号、神经及基于LLM的基线，同时生成可解释的带权规则集成。通过用梯度引导的LLM假设生成替代组合搜索，grasp在不牺牲符号输出透明度的同时保留了提升方法的保证。
英文摘要：
Declarative logic programs offer a powerful and interpretable abstraction for encoding relational structure and neurosymbolic reasoning, by expressing dependencies as weighted compositional rules. However, inducing them from data remains fundamentally hard, bottlenecked by the combinatorial explosion of symbolic search spaces. LLMs have recently emerged as powerful hypothesis generators, but when used in isolation, they lack the capacity to do systematic inductive reasoning needed to reliably synthesize valid programs that fit complex relational distributions. We introduce grasp (Gradient-boosted Synthesis of Probabilistic logic programs), a neurosymbolic framework that casts relational structure learning as functional gradient boosting in which the weak learner is a first-order rule and the intractable inner search is delegated to an LLM proposal oracle. We evaluate grasp on four relational benchmarks spanning molecular toxicity prediction (Tox21), mutagenesis, and citation matching (Cora), and show that it improves over purely symbolic, neural, and LLM-based baselines, while producing interpretable weighted rule ensembles. By replacing combinatorial search with gradient-guided LLM hypothesis generation, grasp retains boosting guarantees without sacrificing the transparency of symbolic outputs.
