LRAS: Advanced Legal Reasoning with Agentic Search
Paper: LRAS: Advanced Legal Reasoning with Agentic Search Authors: Yujin Zhou, Chuxue Cao, Jinluan Yang, Lijun Wu, Conghui He, Sirui Han, et al. (7 authors) Published: 2026-01-12 Link: https://arxiv.org/abs/2601.07296 Category: Agentic Search / Legal AI
Background and Motivation
While Large Reasoning Models (LRMs) have demonstrated exceptional logical capabilities in mathematical domains, their application to the legal field remains hindered by strict requirements for procedural rigor and adherence to legal logic. Existing legal LLMs rely on "closed-loop reasoning" derived solely from internal parametric knowledge. They frequently lack self-awareness of their knowledge boundaries, leading to confident yet incorrect conclusions.
Core Contribution
The paper presents LRAS, the first framework designed to transition legal LLMs from static and parametric *closed-loop thinking* to dynamic and interactive Active Inquiry. It integrates two training techniques:
1. Introspective Imitation Learning — teaching models to recognize the boundaries of their own knowledge; 2. Difficulty-aware Reinforcement Learning — enabling models to handle the complexity of legal reasoning and decide when to invoke search.
Results
- LRAS outperforms state-of-the-art baselines by 8.2–32%.
- The most substantial gains appear in tasks requiring deep reasoning with reliable knowledge.
- The authors state they will release data and models for further exploration.
- Architecture: Agentic paradigms turn retrieval frequency and strategy themselves into learnable decisions, rather than fixed pipeline steps.
- Evaluation: For legal and knowledge-intensive tasks, process metrics (reasoning-chain integrity, citation accuracy) matter as much as end-task accuracy.
- Reliability: Self-awareness of knowledge boundaries — knowing when to search — is a key lever against hallucination in high-stakes domains.
- A Systematic Framework for Enterprise Knowledge Retrieval
- Synergizing RAG and Reasoning: A Systematic Review
- AceSearcher: Bootstrapping Reasoning and Search for LLMs via Reinforcement Learning
- Agentic Information Retrieval
Original Abstract
> While Large Reasoning Models (LRMs) have demonstrated exceptional logical capabilities in mathematical domains, their application to the legal field remains hindered by the strict requirements for procedural rigor and adherence to legal logic. Existing legal LLMs, which rely on "closed-loop reasoning" derived solely from internal parametric knowledge, frequently suffer from lack of self-awareness regarding their knowledge boundaries, leading to confident yet incorrect conclusions. To address this challenge, we present Legal Reasoning with Agentic Search (LRAS), the first framework designed to transition legal LLMs from static and parametric "closed-loop thinking" to dynamic and interactive "Active Inquiry". By integrating Introspective Imitation Learning and Difficulty-aware Reinforcement Learning, LRAS enables LRMs to identify knowledge boundaries and handle legal reasoning complexity. Empirical results demonstrate that LRAS outperforms state-of-the-art baselines by 8.2-32%, with the most substantial gains observed in tasks requiring deep reasoning with reliable knowledge. We will release our data and models for further exploration soon.