# Generative Exaggeration in LLM Social Agents: Consistency, Bias, and Toxicity

arXiv: https://arxiv.org/abs/2507.00657 · HTML: https://mlsanity.com/paper/2507.00657/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Placed in Social simulation and multi-agent debate, one of 16 directions in LLM agents and tool use.
- Builds on 31 of its 60 references that are in the map.
- First posted on arXiv 1 Jul 2025.

## Placed in

- LLM agents and tool use / [Social simulation and multi-agent debate](https://mlsanity.com/agents/social-simulation-and-multi-agent-debate/)

## Builds on (most cited first, 30 of 31)

- [Generative Agents: Interactive Simulacra of Human Behavior](https://mlsanity.com/paper/2304.03442/) (2023)
- [HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face](https://mlsanity.com/paper/2303.17580/) (2023)
- [ChemCrow: Augmenting large-language models with chemistry tools](https://mlsanity.com/paper/2304.05376/) (2023)
- [Social Simulacra: Creating Populated Prototypes for Social Computing Systems](https://mlsanity.com/paper/2208.04024/) (2022)
- [AgentSociety: Large-Scale Simulation of LLM-Driven Generative Agents Advances Understanding of Human Behaviors and Society](https://mlsanity.com/paper/2502.08691/) (2025)
- [A New Generation of Perspective API: Efficient Multilingual Character-level Transformers](https://mlsanity.com/paper/2202.11176/) (2022)
- [Political Compass or Spinning Arrow? Towards More Meaningful Evaluations for Values and Opinions in Large Language Models](https://mlsanity.com/paper/2402.16786/) (2024)
- [Generative Language Models Exhibit Social Identity Biases](https://mlsanity.com/paper/2310.15819/) (2023)
- [Systematic Biases in LLM Simulations of Debates](https://mlsanity.com/paper/2402.04049/) (2024)
- [CoMPosT: Characterizing and Evaluating Caricature in LLM Simulations](https://mlsanity.com/paper/2310.11501/) (2023)
- [The Political Preferences of LLMs](https://mlsanity.com/paper/2402.01789/) (2024)
- [LLM Generated Persona is a Promise with a Catch](https://mlsanity.com/paper/2503.16527/) (2025)
- [Evaluating Large Language Model Biases in Persona-Steered Generation](https://mlsanity.com/paper/2405.20253/) (2024)
- [Large Language Models Reflect the Ideology of their Creators](https://mlsanity.com/paper/2410.18417/) (2024)
- [Y Social: an LLM-powered Social Media Digital Twin](https://mlsanity.com/paper/2408.00818/) (2024)
- [How Susceptible are Large Language Models to Ideological Manipulation?](https://mlsanity.com/paper/2402.11725/) (2024)
- [Emergence of polarization in networks of large language model agents](https://mlsanity.com/paper/2501.05171/) (2025)
- [The simulation of judgment in LLMs](https://mlsanity.com/paper/2502.04426/) (2025)
- [Unmasking Conversational Bias in AI Multiagent Systems](https://mlsanity.com/paper/2501.14844/) (2025)
- [What Large Language Models Do Not Talk About: An Empirical Study of Moderation and Censorship Practices](https://mlsanity.com/paper/2504.03803/) (2025)
- [Political DEBATE: Efficient Zero-shot and Few-shot Classifiers for Political Text](https://mlsanity.com/paper/2409.02078/) (2024)
- [A Public Dataset Tracking Social Media Discourse about the 2024 U.S. Presidential Election on Twitter/X](https://mlsanity.com/paper/2411.00376/) (2024)
- [Robustness and Confounders in the Demographic Alignment of LLMs with Human Perceptions of Offensiveness](https://mlsanity.com/paper/2411.08977/) (2024)
- [Filter bubbles and affective polarization in user-personalized large language model outputs](https://mlsanity.com/paper/2311.14677/) (2023)
- [Entropy and type-token ratio in gigaword corpora](https://mlsanity.com/paper/2411.10227/) (2024)
- [The Impact of Generative AI on Social Media: An Experimental Study](https://mlsanity.com/paper/2506.14295/) (2025)
- [Engagement-Driven Content Generation with Large Language Models](https://mlsanity.com/paper/2411.13187/) (2024)
- [The Impact of Persona-based Political Perspectives on Hateful Content Detection](https://mlsanity.com/paper/2502.00385/) (2025)
- [SALM: A Multi-Agent Framework for Language Model-Driven Social Network Simulation](https://mlsanity.com/paper/2505.09081/) (2025)
- [Understanding Online Polarization Through Human-Agent Interaction in a Synthetic LLM-Based Social Network](https://mlsanity.com/paper/2506.15866/) (2025)

## Built on by (most cited first, 9 of 9)

- [Feel the Difference? A Comparative Analysis of Emotional Arcs in Real and LLM-Generated CBT Sessions](https://mlsanity.com/paper/2508.20764/) (2025)
- [From the Fluency Fallacy to the Micro-to-Macro Validity Gap: Opportunities and Pitfalls of LLMs in Social Simulation](https://mlsanity.com/paper/2507.19364/) (2025)
- [The Statistical Signature of LLMs](https://mlsanity.com/paper/2602.18152/) (2026)
- [Too Open for Opinion? Embracing Open-Endedness in Large Language Models for Social Simulation](https://mlsanity.com/paper/2510.13884/) (2025)
- [Evaluating the Realism of LLM-powered Social Agents: A Case Study of Reactions to Spanish Online News](https://mlsanity.com/paper/2605.28598/) (2026)
- [Large language models for spreading dynamics in complex systems](https://mlsanity.com/paper/2602.08085/) (2026)
- [Collective cooperation without individual fidelity in LLM agents](https://mlsanity.com/paper/2606.30454/) (2026)
- [Practical Judgment, Virtue, and Intuition in the Use of Opaque AI-Enabled Systems](https://mlsanity.com/paper/2607.12755/) (2026)
- [From Caricature to Collective Artifact: How Persona Construction Drives Polarization in Generative Social Simulations](https://mlsanity.com/paper/2610.16498/) (2026)

## Abstract (arXiv)

We investigate how Large Language Models (LLMs) behave when simulating political discourse on social media. Leveraging 21 million interactions on X during the 2024 U.S. presidential election, we construct LLM agents based on 1,186 real users, prompting them to reply to politically salient tweets under controlled conditions. Agents are initialized either with minimal ideological cues (Zero Shot) or recent tweet history (Few Shot), allowing one-to-one comparisons with human replies. We evaluate three model families (Gemini, Mistral, and DeepSeek) across linguistic style, ideological consistency, and toxicity. We find that richer contextualization improves internal consistency but also amplifies polarization, stylized signals, and harmful language. We observe an emergent distortion that we call "generation exaggeration": a systematic amplification of salient traits beyond empirical baselines. Our analysis shows that LLMs do not emulate users, they reconstruct them. Their outputs, indeed, reflect internal optimization dynamics more than observed behavior, introducing structural biases that compromise their reliability as social proxies. This challenges their use in content moderation, deliberative simulations, and policy modeling.