通过分层评分框架提升多模态大模型推理能力，避免错误引导。
RuCL: Stratified Rubric-Based Curriculum Learning for Multimodal Large Language Model Reasoning
- 设计分层评分机制，按模型能力动态调整监督强度。
- 在多个视觉推理任务上实现60.06%准确率，较基线提升7.83%。
- 适合需要高可靠推理的多模态系统开发者使用。
强化学习结合可验证奖励（RLVR）已成为提升多模态大语言模型（MLLMs）推理能力的主流范式。然而，仅依赖结果监督可能导致奖励劫持，即模型学会虚假推理路径以满足最终答案检查。尽管近年基于评分的方法提供了细粒度监督信号，但其面临实例级生成开销大、所有评分项同等学习导致训练效率低的问题。本文提出分层评分式课程学习（RuCL），将课程学习的核心从数据选择转向奖励设计。RuCL生成具有广泛适用性的通用评分标准，并根据模型能力对评分进行分层。训练中动态调整评分权重，引导模型从基础感知逐步过渡到高级逻辑推理。在多个视觉推理基准上的实验表明，RuCL相较Qwen2.5-VL-7B模型平均提升7.83%，达到60.06%的当前最优准确率。
原文摘要 · Abstract (English)
Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a prevailing paradigm for enhancing reasoning in Multimodal Large Language Models (MLLMs). However, relying solely on outcome supervision risks reward hacking, where models learn spurious reasoning patterns to satisfy final answer checks. While recent rubric-based approaches offer fine-grained supervision signals, they suffer from high computational costs of instance-level generation and inefficient training dynamics caused by treating all rubrics as equally learnable. In this paper, we propose Stratified Rubric-based Curriculum Learning (RuCL), a novel framework that reformulates curriculum learning by shifting the focus from data selection to reward design. RuCL generates generalized rubrics for broad applicability and stratifies them based on the model's competence. By dynamically adjusting rubric weights during training, RuCL guides the model from mastering foundational perception to tackling advanced logical reasoning. Extensive experiments on various visual reasoning benchmarks show that RuCL yields a remarkable +7.83% average improvement over the Qwen2.5-VL-7B model, achieving a state-of-the-art accuracy of 60.06%.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。