Towards Real-world Human Behavior Simulation: Benchmarking Large Language Models on Long-horizon, Cross-scenario, Heterogeneous Behavior Traces

2026-04-09

AI Summary

Current large language models (LLMs) struggle to capture the long-term, cross-context, and heterogeneous complexity of real-world human behavior due to reliance on isolated scenarios, constrained action spaces, or synthetic data. This work proposes OmniBehavior—the first high-fidelity behavioral simulation benchmark constructed from authentic user logs—integrating longitudinal, multi-context, and heterogeneous behavioral trajectories to systematically evaluate LLMs’ capacity for behavioral modeling. Experimental results reveal that even with extended context windows, models consistently fail to reproduce complex behaviors, exhibiting structural limitations such as “tunnel vision,” personality homogenization, a bias toward an “optimistically average” persona, and utopian distortions. These findings underscore a fundamental inadequacy in existing approaches to cross-context causal modeling of human behavior.

Abstract

The emergence of Large Language Models (LLMs) has illuminated the potential for a general-purpose user simulator. However, existing benchmarks remain constrained to isolated scenarios, narrow action spaces, or synthetic data, failing to capture the holistic nature of authentic human behavior. To bridge this gap, we introduce OmniBehavior, the first user simulation benchmark constructed entirely from real-world data, integrating long-horizon, cross-scenario, and heterogeneous behavioral patterns into a unified framework. Based on this benchmark, we first provide empirical evidence that previous datasets with isolated scenarios suffer from tunnel vision, whereas real-world decision-making relies on long-term, cross-scenario causal chains. Extensive evaluations of state-of-the-art LLMs reveal that current models struggle to accurately simulate these complex behaviors, with performance plateauing even as context windows expand. Crucially, a systematic comparison between simulated and authentic behaviors uncovers a fundamental structural bias: LLMs tend to converge toward a positive average person, exhibiting hyper-activity, persona homogenization, and a Utopian bias. This results in the loss of individual differences and long-tail behaviors, highlighting critical directions for future high-fidelity simulation research.

Problem

• human behavior simulation
• large language models
• real-world data
• long-horizon behavior
• cross-scenario behavior

Innovation

• OmniBehavior
• long-horizon behavior
• cross-scenario simulation
• heterogeneous behavior traces
• structural bias in LLMs