标签
This paper presents EVIL-Detect, a multi-signal ensemble framework with conflict-aware fusion for detecting LLM-generated, refined, and human-written Chinese text in the NLPCC 2026 Shared Task 6, achieving first place with a macro-F1 of 0.8888.
本文提出了HNR-DAC，一种用于引用论文科学声明验证的两阶段框架，结合了难负样本重排序与分布对齐分类。该方法在NLPCC 2026 Task 10 Track 2上表现优异，以最高的Macro-F1位列排行榜第三名。