Point of Order: Action-Aware LLM Persona Modeling for Realistic Civic Simulation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Merrill, Scott, Srivastava, Shashank
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915860776484864
author Merrill, Scott
Srivastava, Shashank
author_facet Merrill, Scott
Srivastava, Shashank
contents Large language models offer opportunities to simulate multi-party deliberation, but realistic modeling remains limited by a lack of speaker-attributed data. Transcripts produced via automatic speech recognition (ASR) assign anonymous speaker labels (e.g., Speaker_1), preventing models from capturing consistent human behavior. This work introduces a reproducible pipeline to transform public Zoom recordings into speaker-attributed transcripts with metadata like persona profiles and pragmatic action tags (e.g., [propose_motion]). We release three local government deliberation datasets: Appellate Court hearings, School Board meetings, and Municipal Council sessions. Fine-tuning LLMs to model specific participants using this "action-aware" data produces a 67% reduction in perplexity and nearly doubles classifier-based performance metrics for speaker fidelity and realism. Turing-style human evaluations show our simulations are often indistinguishable from real deliberations, providing a practical and scalable method for complex realistic civic simulations.
format Preprint
id arxiv_https___arxiv_org_abs_2511_17813
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Point of Order: Action-Aware LLM Persona Modeling for Realistic Civic Simulation
Merrill, Scott
Srivastava, Shashank
Computation and Language
Artificial Intelligence
Machine Learning
Sound
I.2.7; I.2.6
Large language models offer opportunities to simulate multi-party deliberation, but realistic modeling remains limited by a lack of speaker-attributed data. Transcripts produced via automatic speech recognition (ASR) assign anonymous speaker labels (e.g., Speaker_1), preventing models from capturing consistent human behavior. This work introduces a reproducible pipeline to transform public Zoom recordings into speaker-attributed transcripts with metadata like persona profiles and pragmatic action tags (e.g., [propose_motion]). We release three local government deliberation datasets: Appellate Court hearings, School Board meetings, and Municipal Council sessions. Fine-tuning LLMs to model specific participants using this "action-aware" data produces a 67% reduction in perplexity and nearly doubles classifier-based performance metrics for speaker fidelity and realism. Turing-style human evaluations show our simulations are often indistinguishable from real deliberations, providing a practical and scalable method for complex realistic civic simulations.
title Point of Order: Action-Aware LLM Persona Modeling for Realistic Civic Simulation
topic Computation and Language
Artificial Intelligence
Machine Learning
Sound
I.2.7; I.2.6
url https://arxiv.org/abs/2511.17813