DocThinker: Explainable Multimodal Large Language Models with Rule-based Reinforcement Learning for Document Understanding
This addresses the need for reliable and trustworthy document analysis in high-stakes domains like legal, financial, and medical fields, offering an incremental improvement over existing methods by enhancing adaptability and transparency.
The paper tackles the black-box reasoning problem in Multimodal Large Language Models (MLLMs) for document understanding by proposing DocThinker, a rule-based Reinforcement Learning framework that dynamically refines reasoning strategies, resulting in improved generalization and more explainable outputs across multiple benchmarks.
Multimodal Large Language Models (MLLMs) have demonstrated remarkable capabilities in document understanding. However, their reasoning processes remain largely black-box, making it difficult to ensure reliability and trustworthiness, especially in high-stakes domains such as legal, financial, and medical document analysis. Existing methods use fixed Chain-of-Thought (CoT) reasoning with supervised fine-tuning (SFT) but suffer from catastrophic forgetting, poor adaptability, and limited generalization across domain tasks. In this paper, we propose DocThinker, a rule-based Reinforcement Learning (RL) framework for dynamic inference-time reasoning. Instead of relying on static CoT templates, DocThinker autonomously refines reasoning strategies via policy learning, generating explainable intermediate results, including structured reasoning processes, rephrased questions, regions of interest (RoI) supporting the answer, and the final answer. By integrating multi-objective rule-based rewards and KL-constrained optimization, our method mitigates catastrophic forgetting and enhances both adaptability and transparency. Extensive experiments on multiple benchmarks demonstrate that DocThinker significantly improves generalization while producing more explainable and human-understandable reasoning steps. Our findings highlight RL as a powerful alternative for enhancing explainability and adaptability in MLLM-based document understanding. Code will be available at https://github.com/wenwenyu/DocThinker.