Exploring Lightweight Large Language Models for Court View Generation

2026-05-15

AI Summary

This study addresses the lack of efficient, lightweight large language model (LLM) solutions for court view generation (CVG) and charge prediction in resource-constrained criminal justice settings. It presents the first systematic evaluation of LLMs with fewer than 2 billion parameters on the CVG task and their downstream impact on charge prediction accuracy. The work introduces a novel paradigm that first generates court views and then predicts charges, and contributes CVGEvalKit—an open-source, unified evaluation toolkit encompassing three mainstream model architectures and multiple public datasets. Experimental results demonstrate that lightweight LLMs achieve strong performance on CVG, significantly improve charge prediction accuracy, and strike an effective balance between model scale and performance, thereby validating their practical potential in judicial AI applications.

Abstract

Criminal Court View Generation (CVG) is a critical task in Legal Artificial Intelligence (Legal AI), involving the generation of court view based on case facts. In this work, we systematically explore the capabilities of lightweight (smaller than 2B) large language models (LLMs) in CVG and their impact on charge prediction. Our study addresses four key questions: (1) how does different architecture of LLMs affect the CVG quality and charge prediction. (2) how does LLMs size contribute to the performance, (3) how do lightweight LLMs compare with Deep Neural Networks (DNNs) in these tasks, and (4) how does predicting charge by court view generation first compare with predicting it directly. Additionally, we also develop CVGEvalKit, an evaluation framework including three public available datasets for CVG tasks, as well as predicting their charges. Comprehensive experiments are conducted on this framework, where models are trained on a mixed training set and evaluated on each dataset's test set. Experimental results provide new insights into the trade-offs between model architecture, model size, and the influence between different tasks, highlighting the potential of lightweight LLMs in judicial AI applications. The source code is anonymously available at \url{https://github.com/ZhitianHou/CVGEvalKit}

Problem

• Court View Generation
• Lightweight LLMs
• Legal AI
• Charge Prediction
• Model Architecture

Innovation

• Lightweight LLMs
• Court View Generation
• Legal AI
• Charge Prediction
• CVGEvalKit