一作ACL 2026Poster· Natural Language Processing
Beyond Pedagogical Principles: Multi-Horizon Preference Optimization for Efficient Socratic Tutoring
Xin Shi · Chao Zhang · Yifan Zhu · Xueqiao Zhang · Yawei Luo
摘要
The development of LLM-based tutor agents faces challenges in simultaneously ensuring adherence to pedagogical principles and achieving optimal pedagogical effectiveness, particularly in dynamic, multi-turn interactions. Existing methods are often constrained by static data or sparse reward signa...