GenSim: A General Social Simulation Platform with Large Language Model based Agents

Jiakai Tang,Heyang Gao,Xuchen Pan,Lei Wang,Haoran Tan,Dawei Gao,Yushuo Chen,Xu Chen,Yankai Lin,Yaliang Li,Bolin Ding,Jingren Zhou,Jun Wang,Ji-Rong Wen

2024-10-09

Abstract:With the rapid advancement of large language models (LLMs), recent years have witnessed many promising studies on leveraging LLM-based agents to simulate human social behavior. While prior work has demonstrated significant potential across various domains, much of it has focused on specific scenarios involving a limited number of agents and has lacked the ability to adapt when errors occur during simulation. To overcome these limitations, we propose a novel LLM-agent-based simulation platform called \textit{GenSim}, which: (1) \textbf{Abstracts a set of general functions} to simplify the simulation of customized social scenarios; (2) \textbf{Supports one hundred thousand agents} to better simulate large-scale populations in real-world contexts; (3) \textbf{Incorporates error-correction mechanisms} to ensure more reliable and long-term simulations. To evaluate our platform, we assess both the efficiency of large-scale agent simulations and the effectiveness of the error-correction mechanisms. To our knowledge, GenSim represents an initial step toward a general, large-scale, and correctable social simulation platform based on LLM agents, promising to further advance the field of social science.

Multiagent Systems,Artificial Intelligence

What problem does this paper attempt to address?

The paper attempts to address the following issues:

1. **Limitations of existing social simulation platforms**:
- **Specific scenario limitations**: Existing agent-based simulation platforms based on large language models (LLM) are mostly limited to specific scenarios and cannot be widely applied to various social contexts.
- **Small-scale simulations**: Most existing platforms can only support a limited number of agents, making it difficult to simulate the behavior of large-scale crowds.
- **Lack of error correction mechanisms**: When deviations occur during the simulation process, existing platforms lack effective error correction mechanisms, resulting in poor reliability and long-term stability of the simulation results.

2. **Challenges in social science research**:
- **High cost**: Collecting real human data is expensive.
- **Poor controllability**: It is difficult to control all variables in real-world experiments.
- **Low reproducibility**: Collecting real data and conducting experiments are difficult to replicate at different times and places.

To address the above issues, the paper proposes a new general social simulation platform based on LLM agents—GenSim. The main features of GenSim include:

1. **Generality**: Abstracts a set of general functions to simplify the simulation of customized social scenarios.
2. **Large-scale**: Supports up to 100,000 agents, better simulating the behavior of large-scale crowds in the real world.
3. **Correctability**: Introduces error correction mechanisms to ensure the reliability and long-term stability of the simulation results.

Through these features, GenSim aims to provide a more efficient, reliable, and scalable simulation platform for social science research.