ArgLLM-App: An Interactive System for Argumentative Reasoning with Large Language Models
arXiv:2602.24172v1 Announce Type: new 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.
arXiv:2602.24172v1 Announce Type: new 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.
Executive Summary
The article introduces ArgLLM-App, a web-based system that leverages Large Language Models (LLMs) and computational argumentation for decision-making. The system provides visual explanations and allows human users to interact with and contest the decisions made by the LLM-empowered agents. ArgLLM-App is modular, enabling the integration of trusted external sources, and is publicly available. This innovation has the potential to enhance the explainability and contestability of AI-driven decisions, making them more trustworthy and accountable.
Key Points
- ▸ ArgLLM-App is a web-based system for argumentative reasoning with LLMs
- ▸ The system provides visual explanations and supports human interaction and contestation
- ▸ ArgLLM-App is modular and integrates with trusted external sources
Merits
Enhanced Explainability
ArgLLM-App provides transparent and visual explanations of the decision-making process, increasing trust in AI-driven decisions.
Demerits
Dependence on Data Quality
The accuracy and reliability of ArgLLM-App's decisions are contingent upon the quality and integrity of the data used to train the LLMs and inform the decision-making process.
Expert Commentary
The introduction of ArgLLM-App represents a significant step forward in the development of transparent and accountable AI systems. By providing a platform for human interaction and contestation, ArgLLM-App has the potential to increase trust in AI-driven decisions and promote more informed decision-making. However, it is crucial to address the limitations and challenges associated with the system, such as ensuring the quality and integrity of the data used to train the LLMs. As the use of ArgLLM-App and similar systems becomes more widespread, it is essential to establish robust regulatory frameworks that prioritize transparency, accountability, and human values.
Recommendations
- ✓ Further research should be conducted to evaluate the effectiveness and limitations of ArgLLM-App in various domains and contexts.
- ✓ Regulatory bodies should develop and implement guidelines and standards for the development and deployment of transparent and accountable AI systems, such as ArgLLM-App.