Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

ArgLLM-App: An Interactive System for Argumentative Reasoning with Large Language Models

Adam Dejl, Deniz Gorur, Francesca Toni

TL;DR

This work proposes a web-based system implementing ArgLLM-empowered agents for binary tasks, which supports visualisation of the produced explanations and interaction with human users, allowing them to identify and contest any mistakes in the system's reasoning.

Abstract

Argumentative LLMs (ArgLLMs) are an existing approach leveraging Large Language Models (LLMs) and computational argumentation for decision-making, with the aim of making the resulting decisions faithfully explainable to and contestable by humans. Here we propose a web-based system implementing ArgLLM-empowered agents for binary tasks. ArgLLM-App supports visualisation of the produced explanations and interaction with human users, allowing them to identify and contest any mistakes in the system's reasoning. It is highly modular and enables drawing information from trusted external sources. ArgLLM-App is publicly available at https://argllm.app, with a video demonstration at https://youtu.be/vzwlGOr0sPM.

ArgLLM-App: An Interactive System for Argumentative Reasoning with Large Language Models

TL;DR

This work proposes a web-based system implementing ArgLLM-empowered agents for binary tasks, which supports visualisation of the produced explanations and interaction with human users, allowing them to identify and contest any mistakes in the system's reasoning.

Abstract

Argumentative LLMs (ArgLLMs) are an existing approach leveraging Large Language Models (LLMs) and computational argumentation for decision-making, with the aim of making the resulting decisions faithfully explainable to and contestable by humans. Here we propose a web-based system implementing ArgLLM-empowered agents for binary tasks. ArgLLM-App supports visualisation of the produced explanations and interaction with human users, allowing them to identify and contest any mistakes in the system's reasoning. It is highly modular and enables drawing information from trusted external sources. ArgLLM-App is publicly available at https://argllm.app, with a video demonstration at https://youtu.be/vzwlGOr0sPM.
Paper Structure (4 sections, 2 figures)

This paper contains 4 sections, 2 figures.

Table of Contents

  1. Introduction
  2. Preliminaries
  3. ArgLLM-App Features
  4. Future Work

Figures (2)

  • Figure 1: Output of ArgLLM-App for claim "Ukraine will join the EU before 2030", with settings as in Figure \ref{['fig:settings']} and document-based QBAF generation (see the PDF document indication at the bottom of the chat). In the QBAF, the bottom left argument has been selected for addition of a supporter at depth 3. The abstract graph at the bottom right shows the QBAF structure.
  • Figure 2: Settings: in addition to the API key for access to the base LLM, the user can configure the gradual semantics, depth and breadth to be used by ArgLLM-App.