# SELF-PERCEPT: Introspection Improves Large Language Models' Detection of Multi-Person Mental Manipulation in Conversations

arXiv: https://arxiv.org/abs/2505.20679 · HTML: https://mlsanity.com/paper/2505.20679/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 4 of its 24 references that are in the map.
- First posted on arXiv 27 May 2025.

## Builds on (most cited first, 4 of 4)

- [MentalManip: A Dataset For Fine-grained Analysis of Mental Manipulation in Conversations](https://mlsanity.com/paper/2405.16584/) (2024)
- [Detecting Conversational Mental Manipulation with Intent-Aware Prompting](https://mlsanity.com/paper/2412.08414/) (2024)
- [Exploring Effective Information Utilization in Multi-Turn Topic-Driven Conversations](https://mlsanity.com/paper/2209.00250/) (2022)
- [A Global-Local Attention Mechanism for Relation Classification](https://mlsanity.com/paper/2407.01424/) (2024)

## Built on by (most cited first, 2 of 2)

- [The Hidden Puppet Master: Predicting Human Belief Change in Manipulative LLM Dialogues](https://mlsanity.com/paper/2603.20907/) (2026)
- [From Pattern Recognizers to Personalized Companions: A Survey of Large Language Models in Mental Health](https://mlsanity.com/paper/2609.25186/) (2026)

## Abstract (arXiv)

Mental manipulation is a subtle yet pervasive form of abuse in interpersonal communication, making its detection critical for safeguarding potential victims. However, due to manipulation's nuanced and context-specific nature, identifying manipulative language in complex, multi-turn, and multi-person conversations remains a significant challenge for large language models (LLMs). To address this gap, we introduce the MultiManip dataset, comprising 220 multi-turn, multi-person dialogues balanced between manipulative and non-manipulative interactions, all drawn from reality shows that mimic real-world scenarios. For manipulative interactions, it includes 11 distinct manipulations depicting real-life scenarios. We conduct extensive evaluations of state-of-the-art LLMs, such as GPT-4o and Llama-3.1-8B, employing various prompting strategies. Despite their capabilities, these models often struggle to detect manipulation effectively. To overcome this limitation, we propose SELF-PERCEPT, a novel, two-stage prompting framework inspired by Self-Perception Theory, demonstrating strong performance in detecting multi-person, multi-turn mental manipulation. Our code and data are publicly available at https://github.com/danushkhanna/self-percept .