arXivDaily arXiv每日学术速递 周一至周五更新
​通知:联邦节假日调休:哥伦布日 / 原住民日(Columbus Day / Indigenous Peoples' Day),今日(13日)无更新。
arXiv 2603.04319cs.CL

AILS-NTUA 在 SemEval-2026 任务 12: 基于图的检索与反思提示的归纳事件推理

AILS-NTUA at SemEval-2026 Task 12: Graph-Based Retrieval and Reflective Prompting for Abductive Event Reasoning

School of Electrical and Computer Engineering, AILS Laboratory(电气与计算机工程学院,AILS实验室);National Technical University of Athens(雅典国家技术大学)

Nikolas Karafyllis, Maria Lymperaiou, Giorgos Filandrianos, Athanasios Voulodimos, Giorgos Stamou

AI总结:

AILS-NTUA 提出基于图检索与反思提示的三阶段系统,在 SemEval-2026 任务 12 中实现 0.95 准确率,揭示多标签因果推理中的系统性失败模式。

摘要:

我们提出了一种在 SemEval 2026 任务 12: 归纳事件推理中获胜的三阶段系统,结合了基于图的检索、由大语言模型驱动的归纳推理以及通过反思提示进化优化的提示设计,以及事后一致性强制执行;我们的系统在评估阶段排行榜上排名第一,准确率为 0.95。跨模型误差分析显示 14 个模型(7 个家族)共有的三个归纳偏见:因果链不完整、近因偏好和显著性偏见,其跨家族收敛(51% 原因计数减少)表明在多标签因果推理中存在系统性而非模型特定的失败模式。

英文摘要:

We present a winning three-stage system for SemEval 2026 Task~12: Abductive Event Reasoning that combines graph-based retrieval, LLM-driven abductive reasoning with prompt design optimized through reflective prompt evolution, and post-hoc consistency enforcement; our system ranks first on the evaluation-phase leaderboard with an accuracy score of 0.95. Cross-model error analysis across 14 models (7~families) reveals three shared inductive biases: causal chain incompleteness, proximate cause preference, and salience bias, whose cross-family convergence (51\% cause-count reduction) indicates systematic rather than model-specific failure modes in multi-label causal reasoning.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑