arXiv AI By Xiaokun Wang, Siyu Song, Wentao Liu, Xiaodong Zou

Beyond Direct Answering: Aligning Educational LLMs as Socratic Guides via Heuristic Reinforcement Learning

Read the original on arXiv AI →

arXiv:2607. 22996v1 Announce Type: cross Abstract: Large language models (LLMs) deployed in educational settings often behave as direct answerers: they disclose target concepts in the opening turn instead of guiding students through progressive inquiry, as Socratic pedagogy prescribes.

Summary generated by The Flow from the publisher's feed. The full article lives at arXiv AI.

arXiv AI
Jul 23

Mitigating Scaffolding Collapse in Socratic Tutors via Representation Alignment

arXiv:2607. 19371v1 Announce Type: new Abstract: Large language model (LLM)-based Socratic tutors increasingly guide students through multi-turn questioning, but they can suffer from scaffolding collapse: under sustained student pressure, a tutor gradually abandons guided inquiry and reveals solutions directly.

By Jing Shao, Qifeng Wu, Hanyu Zhang, Sixia Sun, Jun Zhuang