Benchmarking Argumentative Behaviour of LLMs: A Study of Defences Against Character Attacks
Assessing how argumentative AI models handle ad hominem attacks compared to human counterparts.
Large Language Models (LLMs) are increasingly deployed as argumentative agents in persuasive dialogues, necessitating rigorous evaluation of their debating competence relative to human interlocutors. In this study, we focus on character attacks (ad hominem arguments), traditional...
ArXiv CS.CL
Source
Generated insight by OnAirToday AI
As LLMs take on more complex roles in persuasive dialogue, understanding their susceptibility to ad hominem attacks is crucial for building trustworthy AI agents. This research provides a necessary framework to ensure that argumentative AI can maintain logical integrity rather than succumbing to common logical fallacies.
Comments
Keep comments relevant and respectful. No links, advertising, self-promotion, or spam.