发表机构
Shandong University; Boston University; North China Electric Power University; National University of Singapore; Joint SDU-NTU Centre for Artificial Intelligence Research (C-FAIR), Shandong University; Rizhao Steel Holding Group Co., Ltd.（山东大学; 波士顿大学; 华北电力大学; 新加坡国立大学; 山东大学-南洋理工大学人工智能联合研究中心（C-FAIR）; 日照钢铁控股集团有限公司）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究针对长周期制造检查日志的早期风险筛查需求，提出ConMem贡献感知记忆框架，通过Shapley风格评估保留高价值证据，在真实数据集上实现76.0% QA准确率，大幅减少输入token数与响应时间，可有效保留弱早期信号并提供预警。
AI 中文摘要
长周期钢铁设备检查需要对多次检查周期中积累的异构记录进行推理。现有检索增强生成系统将历史日志视为静态语料库，保留记录时不评估其诊断价值，无法报告早期风险。为此，我们提出ConMem，一种面向大语言模型（LLM）辅助设备检查的贡献感知记忆框架，支持人在回路的早期风险筛查系统。具体而言，ConMem首先将检查日志分割为功能证据单元，然后通过Shapley风格估计每个记忆单元对下游诊断的贡献，最后在受限的内存预算下保留高价值证据。实验中，我们在真实数据集上评估ConMem，其问答（QA）准确率达76.0%，超过最强直接可比基线。相较于朴素的8K上下文LLM基线，它将平均输入token数减少88.2%，响应时间缩短86.6%。消融研究还表明，功能角色感知分割和基于贡献的估值有助于优先处理弱退化信号，以进行针对性现场检查。实际部署进一步证实，ConMem在三个检查周期中保留了弱早期信号，为现场检查员提供了针对密封磨损的早期预警。
英文摘要
Long-horizon steel-equipment inspection requires reasoning over heterogeneous records accumulated across repeated inspection cycles. Existing retrieval-augmented generation systems treat historical logs as a static corpus and retain records without estimating their diagnostic value, failing to report early risk. To this end, we propose ConMem, a contribution-aware memory framework for LLM-assisted equipment inspection, supporting a human-in-the-loop early-risk screening system. Specifically, our ConMem first segments inspection logs into functional evidence units, then estimates each memory unit's contribution to downstream diagnosis through a Shapley-style estimation, and finally retains high-value evidence under a constrained memory budget. In experiments, we evaluate ConMem on real-world dataset and ConMem achieves 76.0% QA accuracy, exceeding the strongest directly comparable baseline. Relative to the naive 8K-context LLM baselines, it reduces the average number of input tokens by 88.2% and response time by 86.6%. Ablation studies also show that the functional-role-aware segmentation and contribution-based valuation are helping prioritize weak degradation signals for targeted field inspection. Practical deployments further confirm that ConMem retains the weak early signal across three inspection cycles, providing an early-stage seal-wear alert targeted for on-site inspectors.
