arXiv:2505.22961v3 Announce Type: replace Abstract: Large language models (LLMs) have shown promising potential in persuasion, but existing works on training LLM persuaders are still preliminary. Notably, while humans are skilled in modeling their opponent's thoughts and opinions proactively and dynamically, current LLMs struggle with such Theory of Mind (ToM) reasoning, resulting in limited diversity and opponent awareness. To address this limitation, we introduce Theory of Mind Augmented Persuader (ToMAP), a novel approach for building more flexible persuader agents by incorporating two theo
The proliferation of LLMs creates a pressing need to enhance their sophisticated interaction capabilities, especially in complex tasks like persuasion, driving research into advanced reasoning like Theory of Mind.
Improving LLM persuasion through Theory of Mind makes AI agents significantly more versatile and effective in dynamic human-like interactions, impacting areas from customer service to strategic negotiation.
LLMs can now be trained to anticipate and model the thoughts and opinions of their interlocutors, leading to more adaptive and diverse persuasive strategies rather than static responses.
- · AI developers
- · Businesses using AI for customer interaction
- · Researchers in cognitive AI
- · LLMs without advanced reasoning capabilities
- · Static AI interaction models
LLMs become more effective in roles requiring negotiation and influence, potentially streamlining many white-collar workflows.
The development accelerates the deployment of sophisticated AI agents across various sectors, increasing efficiency and potentially displacing more routine human tasks.
Enhanced AI persuasion capabilities could raise ethical concerns about manipulation and the transparency of AI intent in human-AI interactions.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.CL