REAP：面向基于大语言模型的闭卷知识库构建的关系感知引出与解析
REAP: Relation-Aware Elicitation and Parsing for Closed-Book Knowledge Base Construction from LLMs
浏览论文内容
中文总结 AI 辅助
该研究提出REAP系统，基于Mistral-Small-24B-Instruct-2501模型，结合结构化思维链推理等技术，在2026年AKBC共享任务闭卷知识库构建任务中取得宏F1值0.62，相关代码已公开。
中文摘要 AI 辅助
我们提出了REAP系统，用于2026年AKBC共享任务中闭卷场景下从语言模型构建知识库的任务，该任务要求使用最多320亿参数的模型且不进行模型微调。我们的系统结合结构化思维链推理、关系特定查询策略和基于推理的空集门来引出参数化知识，随后直接提取为有效的JSON数组。在测试集上，基于Mistral-Small-24B-Instruct-2501模型构建的该系统取得了宏F1值0.62，在countryLandBordersCountry（F1=0.95）、companyTradesAtStockExchange（F1=0.73）和hasArea（F1=0.77）任务上表现尤为突出。我们的代码可在指定的此URL公开获取。
英文摘要
We present the REAP system for the AKBC Shared Task 2026 on constructing knowledge bases from language models in a closed-book setting, subject to a budget of at most 32B parameters and no model fine-tuning. Our system combines structured chain-of-thought reasoning, relation-specific query strategies, and a reasoning-based empty-set gate to elicit parametric knowledge, followed by direct extraction into valid JSON arrays. On the test set, the system, built on the Mistral-Small-24B-Instruct-2501 model, achieves a macro-F1 score of 0.62, with particularly strong results on countryLandBordersCountry (F1 = 0.95), companyTradesAtStockExchange (F1 = 0.73), and hasArea (F1 = 0.77). Our code is publicly available at https://github.com/yammdd/AKBC-Shared-Task-2026.
发表机构
- VNU University of Engineering and Technology（VNU工程技术大学）
机构由 AI 辅助整理，请以论文原文为准。
