Minwoo Kang

Humans choose how to act via a nuanced process that depends both on rational deliberation and also identity and contextual factors. In this work, we study how large language models (LLMs) might simulate human action in the context of social dilemma games. While prior work has focused on ``steering'' (weak binding) of agent models to simulate personas, we analyze here how deep binding of base models with extended backstories leads to more faithful replication of identity-based behaviors. Our study has these findings: simulation fidelity vs human studies is improved by conditioning base LMs with a rich context of narrative identities and checking consistency using instruction-tuned models. We show that LLMs can also model contextual factors such as time (year that a study was performed) and Question Framing effects. LLMs, therefore allow us to explore the details that affect human studies but which are often omitted from experiment descriptions, and which prevent accurate replication.

Large Language Model (LLM)-based simulators often reproduce observable actions but fail to capture the underlying reasoning behind them. In education, where student simulation is increasingly used for various applications such as evaluating tutoring systems, this gap is especially pronounced. Two students may submit identical submissions for entirely different reasons. We present INTERNAL STUDENT DIALOGUE (INSIDE), a student modeling framework that fine-tunes LLMs not only to act like students but also to think like them. INSIDE generates internal dialogue grounded in Bloom’s Taxonomy across cognitive, affective, and action dimensions, and fine-tunes models on paired think traces and actions. We baseline against different prompting frameworks and evaluate on two axes: fidelity of simulated actions and quality of generated internal dialogue. Our evaluations show that INSIDE improves simulation fidelity in both action fidelity, matching code generation of real students, and reasoning alignment, achieving the highest alignment across models up to 57.9%.