arXiv cs.AI· Haotian Chen, Shuaicheng Niu, Haocong Rao, Kaisong Song, Jun Lin, Lizhen Cui, Zhiqi Shen, Yonghui Xu·· 10 天前AI 评分28
面向长程法律推理的测试时智能体演化
Test-Time Agent Evolution for Long-Horizon Legal Reasoning
AI 导读
研究提出一种免训练的测试时智能体自适应方法，通过测试时记忆演化从既往案例中检索可复用经验并适配当前事实与程序语境，同时以评分标准对齐协作机制校验和修正各角色行动。在 J1-EVAL 和 LegalWorld 上对五个基座模型的实验显示，该方法相较代表性推理与智能体基线取得一致提升，且交互与计算成本合理。
正文 · 原文
Abstract:Legal intelligence aims to support reliable decision-making across long-horizon legal processes involving evolving case states and multiple roles. However, real-world legal deployment exhibits substantial case heterogeneity in facts, evidence, and procedural contexts, exposing the limitations of static agent strategies. Moreover, legal reasoning is inherently interdependent across roles and procedural stages, making global reliability fundamentally different from isolated role competence. To address these challenges, we study training-free test-time agent adaptation, where agents continuously exploit deployment-time signals from preceding cases and ongoing interactions without updating model parameters. We propose \method, which introduces \emph{Test-Time Memory Evolution} to retrieve reusable experience from previous cases, adapt it to the current factual and procedural context, and consolidate accumulated experience for subsequent decision-making. Further, \emph{Rubric-Aligned Collaboration} verifies and revises role-specific actions according to behavioral and procedural requirements, enabling coordinated decision-making across roles and stages. Extensive experiments on J1-EVAL and LegalWorld across five backbone models demonstrate consistent improvements over representative reasoning and agent baselines with reasonable interaction and computational costs. Ablation and case studies further show that the two components provide complementary benefits in experience adaptation and cross-role coordination, improving the reliability and efficiency of long-horizon legal reasoning.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.08138 [cs.AI] |
| (or arXiv:2610.08138v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.08138 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Haotian Chen [view email]
[v1]
Tue, 6 Oct 2026 10:53:25 UTC (893 KB)
来源：arXiv cs.AI · arxiv.org