Pedagogy-R1: Pedagogically-Aligned Reasoning Model with Balanced Educational Benchmark

2025-05-24

AI Summary

Large reasoning models (LRMs) exhibit weak pedagogical coherence, deficient knowledge-transmission logic, and inadequate simulation of teacher behaviors in educational settings. Method: This paper proposes a teaching-aligned distillation fine-tuning paradigm: (1) constructing WBEB—the first multi-dimensional benchmark for evaluating educational capabilities; (2) designing Chain-of-Pedagogy (CoP), a structured prompting strategy that emulates instructional reasoning; and (3) integrating model distillation with instruction tuning to explicitly model pedagogical behaviors. The approach combines quantitative evaluation and qualitative analysis across five core educational tasks. Results: Experiments demonstrate significant improvements in teaching consistency, decision traceability, and reasoning plausibility. For the first time, this work systematically characterizes the strengths and critical limitations of LRMs in pedagogical competence, establishing both theoretical foundations and actionable technical pathways for trustworthy adaptation of large models to education.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Computer Vision: Large Vision Models

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Recent advances in large reasoning models (LRMs) show strong performance in structured domains such as mathematics and programming; however, they often lack pedagogical coherence and realistic teaching behaviors. To bridge this gap, we introduce Pedagogy-R1, a framework that adapts LRMs for classroom use through three innovations: (1) a distillation-based pipeline that filters and refines model outputs for instruction-tuning, (2) the Well-balanced Educational Benchmark (WBEB), which evaluates performance across subject knowledge, pedagogical knowledge, tracing, essay scoring, and teacher decision-making, and (3) a Chain-of-Pedagogy (CoP) prompting strategy for generating and eliciting teacher-style reasoning. Our mixed-method evaluation combines quantitative metrics with qualitative analysis, providing the first systematic assessment of LRMs' pedagogical strengths and limitations.

Problem

• LRMs lack pedagogical coherence in teaching
• Need balanced evaluation across educational metrics
• Require teacher-style reasoning generation methods

Innovation

• Distillation-based pipeline for instruction-tuning
• Well-balanced Educational Benchmark for evaluation
• Chain-of-Pedagogy prompting for teacher-style reasoning