EduFlow: Advancing MLLMs' Problem-Solving Proficiency through Multi-Stage, Multi-Perspective Critique

TL;DR

EduFlow is a comprehensive framework that improves multimodal large language models' scientific reasoning by integrating multi-stage critique, self-reflection, and curriculum learning, leading to more coherent and reliable problem-solving.

Contribution

The paper introduces EduFlow, a novel end-to-end educational reasoning framework with EduPRM and EduMCTS, enabling dynamic multi-stage critique and iterative refinement in scientific reasoning tasks.

Findings

Enhanced reasoning consistency and coherence in MLLMs.

Improved problem-solving accuracy on scientific tasks.

Constructed a large dataset of educational reasoning trajectories.

Abstract

Multimodal large language models (MLLMs) still perform poorly on scientific tasks, particularly those requiring multi-step and interpretable reasoning. Their limitations include insufficient scientific reasoning patterns, lack of global coherence in multi-step inference, and the absence of reflective self-correction, making them unreliable in structured scientific contexts. We introduce EduFlow, the first end-to-end framework that covers the full pipeline of educational scientific reasoning, including data selection, MCTS-based trajectory construction, model training, and output optimization. At its core is EduPRM, a process-aware reward model that critiques reasoning steps with tags and justifications. EduPRM is trained via curriculum learning on three complementary supervision sources: MCTS-guided trajectories, error-injected critiques, and teacher-student dialogues, enabling dynamic adaptation to multi-stage problem solving and iterative refinement during inference. We further propose EduMCTS, a domain-adapted search framework that introduces bootstrapping actions specifically designed for educational reasoning, such as a self-reflection mechanism that promotes reflective error correction. It further leverages EduPRM's fine-grained feedback to guide the search toward higher-quality reasoning trajectories. By applying self-consistency and rejection sampling, we constructed EduMCTS-160K, a large-scale dataset of educational reasoning trajectories. Extensive experiments demonstrate that EduFlow enhances reasoning consistency and coherence. Code, data, and models will be released.