Using LLMs to Model Arguments in U.S. Supreme Court Briefs: Preliminary Report

Authors

• Heng Zheng, University of KentuckyFollow
• Dexter Williams, University of Wisconsin–MadisonFollow
• Bertram Ludäscher, University of Illinois Urbana-ChampaignFollow

Abstract

Supreme Court briefs can impact the decision-making of Supreme Court cases. We show how to represent the briefs for a U.S. Supreme Court case as a formal argument model that relies on the briefs’ existing structure. We explore how to generate argument models with large language models (LLMs) by using them to identify attacks between arguments.

Document Type

Article

Publication Date

2025

Notes/Citation Information

Published in the Proceedings of the International Workshop on Translating Natural Legal Language into Formal Representation (NLL2FR 2025) associated with JURIX 2025.

Funding Information

This work was funded in part by the 2025 Cardiff University & University of Illinois System Joint Research and Innovation Seed Grants Program.

Repository Citation

Zheng, Heng; Williams, Dexter; and Ludäscher, Bertram, "Using LLMs to Model Arguments in U.S. Supreme Court Briefs: Preliminary Report" (2025). Information Science Faculty Publications. 104.
https://uknowledge.uky.edu/slis_facpub/104