LLMs as Debate Partners: Utilizing Genetic Algorithms and Adversarial Search for Adaptive Arguments

Prakash Aryan · 2024

Read LLMs as Debate Partners: Utilizing Genetic Algorithms and Adversarial Search for Adaptive Arguments by Prakash Aryan (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-06229.txt.

Description

Prakash Aryan presents DebateBrawl, a platform where a language model debates human users and adapts as it goes. Plain chat models plan strategy poorly, so the system adds two search tools: a genetic algorithm that evolves arguments and an adversarial search that plans moves like a game. The result aims for a sparring partner that stays sharp and on topic.

In 23 debates, the system scored on par with human users, a hair higher on average. Its claims held up better on fact checks than human-only debates. Most users said their debating skills improved, and the paper closes with safeguards for the platform.

Catalog record

CREATORPrakash Aryan
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE40 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-06229

Browse the full library · About the Tower of Babel