SHAPE：统一安全、帮助性和教学法以用于教育LLM
SHAPE: Unifying Safety, Helpfulness and Pedagogy for Educational LLMs
AI总结：
本文提出SHAPE基准，通过知识掌握图统一安全、帮助性和教学行为，改进教育LLM在对抗压力下的安全性和帮助性。
摘要：
大型语言模型（LLMs）在教育场景中已被广泛研究。我们识别出当前教育LLM的关键漏洞——教学性劫持，即学生使用引导回答的提示来获取解决方案而非构建式指令。为实现系统研究，我们通过知识掌握图统一并形式化安全、帮助性和教学行为，并引入SHAPE基准，包含9,087个学生-问题对，用于评估在对抗压力下的教学行为。我们提出了一种图增强的教学流水线，通过查询推断先决概念，识别掌握差距，并通过显式门控在指导和问题解决之间路由生成。在多个LLM上的实验表明，我们的方法在两种教学性劫持设置下显著提高了安全性，同时在相同评估协议下保持了接近天花板的帮助性。我们的代码和数据可在https://github.com/MAPS-research/SHaPE获取。
英文摘要：
Large Language Models (LLMs) have been widely explored in educational scenarios. We identify a critical vulnerability in current educational LLMs, pedagogical jailbreaks, where students use answer-inducing prompts to elicit solutions rather than scaffolded instructions. To enable systematic study, we unify and formalize safe, helpful, and pedagogical behaviors with a knowledge-mastery graph and introduce SHAPE, a benchmark of 9,087 student-question pairs for evaluating tutoring behavior under adversarial pressure. We propose a graph-augmented tutoring pipeline that infers prerequisite concepts from queries, identifies mastery gaps, and routes generation between instructing and problem-solving via explicit gating. Experiments across multiple LLMs show that our method yields significantly improved safety under two pedagogical jailbreak settings, while maintaining near-ceiling helpfulness under the same evaluation protocol. Our code and data are available at https://github.com/MAPS-research/SHaPE
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。
