Dude: A Dual-Detection Multi-Agent System for Paper-Code Discrepancy Detection

Dude：用于论文与代码不一致性检测的双重检测多智能体系统

Abstract: LLM-empowered paper-code discrepancy detection has received growing concern since the scaling of research submissions exceeds the manual review capability. However, the limited context capacity and one-sided discrepancy detection of existing single-agent LLM paradigms lead to an inferior recall performance in detecting discrepancies.

摘要： 随着研究投稿规模的扩大，人工评审能力已难以应对，基于大语言模型（LLM）的论文与代码不一致性检测日益受到关注。然而，现有的单智能体 LLM 范式存在上下文容量有限和不一致性检测片面等问题，导致在检测不一致性时的召回率表现不佳。

In this paper, we propose Dude, the first Dual-Detection Multi-Agent System for paper-code discrepancy detection. We discover that the granularity asymmetry of the paper-language and code-language introduces over-interpretation and over-reporting challenges in a multi-agent system design for discrepancy detection, resulting in increasing false positives.

在本文中，我们提出了 Dude，这是首个用于论文与代码不一致性检测的双重检测多智能体系统。我们发现，论文语言与代码语言之间的粒度不对称性，在用于不一致性检测的多智能体系统设计中引入了过度解读和过度报告的挑战，从而导致误报率增加。

To address this, we propose a granularity-aligned negotiation and a two-stage salience-filtering mechanism in Dude, which effectively prevents agents from falsely reporting discrepancies. Experimental results in real-world paper-code discrepancy datasets showcase Dude’s significant recall and precision improvement by up to 22.8%, increasing F1 score by up to 18.7% compared to baseline methods.

为了解决这一问题，我们在 Dude 中提出了一种粒度对齐协商机制和两阶段显著性过滤机制，有效防止了智能体对不一致性的误报。在真实世界的论文-代码不一致性数据集上的实验结果表明，与基准方法相比，Dude 的召回率和精确率显著提升了高达 22.8%，F1 分数提高了高达 18.7%。

---

Paper Information:

• Authors: Weijie Liu, Running Zhao, Wenhao Yuan, Jinfeng Xu, Zhanfeng Xu, Xiaoxi Zhang, Edith Cheuk-Han Ngai
• arXiv ID: 2609.03416
• Subjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)

论文信息：

• 作者： Weijie Liu, Running Zhao, Wenhao Yuan, Jinfeng Xu, Zhanfeng Xu, Xiaoxi Zhang, Edith Cheuk-Han Ngai
• arXiv ID: 2609.03416
• 学科分类： 人工智能 (cs.AI)；机器学习 (cs.LG)