Legal$Δ$: 通过强化学习与链式思维引导信息增益提升大语言模型的法律推理
Legal$Δ$: Enhancing Legal Reasoning in LLMs via Reinforcement Learning with Chain-of-Thought Guided Information Gain
AI总结：
Legal$Δ$通过强化学习与链式思维引导信息增益提升法律推理的准确性和可解释性。
摘要：
法律人工智能（LegalAI）已在大型语言模型（LLMs）的支持下实现了自动化司法决策的显著进展。然而，现有的法律LLMs仍然难以生成可靠且可解释的推理过程。它们往往倾向于快速思考行为，通过直接生成答案而不进行显式的多步骤推理，限制了其在需要严格论证的复杂法律场景中的有效性。为了解决这一挑战，我们提出了Legal$Δ$，一种通过链式思维引导信息增益来增强法律推理的强化学习框架。在训练过程中，Legal$Δ$采用包含直接答案和推理增强模式的双模式输入设置，并最大化两者之间的信息增益。这促使模型获取有意义的推理模式，而不是生成表面或冗余的解释。Legal$Δ$采用两阶段方法：（1）从强大的大型推理模型（LRM）DeepSeek-R1中蒸馏潜在的推理能力；（2）通过差异比较来优化推理质量，结合一个多维奖励机制，评估结构连贯性和法律领域特定性。在多个法律推理任务上的实验结果表明，Legal$Δ$在准确性和可解释性方面均优于强大的基线模型。它能够持续生成更加稳健和可信的法律判断，而无需依赖标注的偏好数据。所有代码和数据将在https://github.com/NEUIR/LegalDelta上发布。
英文摘要：
Legal Artificial Intelligence (LegalAI) has achieved notable advances in automating judicial decision-making with the support of Large Language Models (LLMs). However, existing legal LLMs still struggle to generate reliable and interpretable reasoning processes. They often default to fast-thinking behavior by producing direct answers without explicit multi-step reasoning, limiting their effectiveness in complex legal scenarios that demand rigorous justification. To address this challenge, we propose Legal$Δ$, a reinforcement learning framework designed to enhance legal reasoning through chain-of-thought guided information gain. During training, Legal$Δ$ employs a dual-mode input setup-comprising direct answer and reasoning-augmented modes-and maximizes the information gain between them. This encourages the model to acquire meaningful reasoning patterns rather than generating superficial or redundant explanations. Legal$Δ$ follows a two-stage approach: (1) distilling latent reasoning capabilities from a powerful Large Reasoning Model (LRM), DeepSeek-R1, and (2) refining reasoning quality via differential comparisons, combined with a multidimensional reward mechanism that assesses both structural coherence and legal-domain specificity. Experimental results on multiple legal reasoning tasks demonstrate that Legal$Δ$ outperforms strong baselines in both accuracy and interpretability. It consistently produces more robust and trustworthy legal judgments without relying on labeled preference data. All code and data will be released at https://github.com/NEUIR/LegalDelta.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
