Decoding the Human Factor: High Fidelity Behavioral Prediction for Strategic Foresight

TL;DR

This work tackles the problem of predicting individual strategic choices in high-stakes contexts by moving beyond prompt-based persona conditioning to a structured, trait-conditioned embedding via the Large Behavioral Model (LBM). Leveraging a psychometric battery of 74 traits and a bank of 55 scenarios, LBM is fine-tuned with LoRA on a Llama backbone to map stable dispositions and situational constraints to discrete actions, achieving measurable gains over an unadapted backbone and competitive performance against frontier baselines when conditioned on Big Five traits. Crucially, performance scales with trait dimensionality, revealing a complexity ceiling for prompting baselines but continued benefit from richer trait profiles for LBM. The results support LBM as a scalable framework for high-fidelity behavioral simulation with applications in strategic foresight, negotiation analysis, and decision support, while highlighting limitations related to sample, ecology, and generalizability and outlining directions for broader, longer-horizon modeling.

Abstract

Predicting human decision-making in high-stakes environments remains a central challenge for artificial intelligence. While large language models (LLMs) demonstrate strong general reasoning, they often struggle to generate consistent, individual-specific behavior, particularly when accurate prediction depends on complex interactions between psychological traits and situational constraints. Prompting-based approaches can be brittle in this setting, exhibiting identity drift and limited ability to leverage increasingly detailed persona descriptions. To address these limitations, we introduce the Large Behavioral Model (LBM), a behavioral foundation model fine-tuned to predict individual strategic choices with high fidelity. LBM shifts from transient persona prompting to behavioral embedding by conditioning on a structured, high-dimensional trait profile derived from a comprehensive psychometric battery. Trained on a proprietary dataset linking stable dispositions, motivational states, and situational constraints to observed choices, LBM learns to map rich psychological profiles to discrete actions across diverse strategic dilemmas. In a held-out scenario evaluation, LBM fine-tuning improves behavioral prediction relative to the unadapted Llama-3.1-8B-Instruct backbone and performs comparably to frontier baselines when conditioned on Big Five traits. Moreover, we find that while prompting-based baselines exhibit a complexity ceiling, LBM continues to benefit from increasingly dense trait profiles, with performance improving as additional trait dimensions are provided. Together, these results establish LBM as a scalable approach for high-fidelity behavioral simulation, enabling applications in strategic foresight, negotiation analysis, cognitive security, and decision support.

Figures (2)

• Figure 1: Comparison of Behavioral Alignment across LLMs. Bars report accuracy, balanced accuracy, and macro-F1 for the base backbone model, the SFT-tuned LBM, and frontier LLM baselines. Error bars indicate 95% confidence intervals computed using participant-level bootstrap resampling. (Rendered from PDF page 11.)
• Figure 2: Effect of trait dimensionality on behavioral prediction. Lines show model performance when conditioned on 5, 10, 20, 40 and 74 traits. (a) Accuracy. (b) Macro-F1. LBM improves with additional traits, while the base model and Claude remain approximately constant; Grok shows a modest upward trend but remains below LBM.