Year: 2026
Type: conference-paper
Abstract: We study self-modeling: an LLM's ability to answer questions about its own behavior.We focus on verifiable behavioral questions, such as whether a prompt edit would change the model's final answer.To measure this capability, we introduce a benchmark that tests diverse types of self-modeling questions.Current models show non-trivial but limited selfmodeling skill, and make systematic mistakes on simple counterfactual questions about their own behavior.To improve self-modeling skill, we develop a scalable synthetic-data pipeline that produces self-modeling training data, and show that reinforcement-learning can improve aggregate self-modeling skill across three opensource model families with some transfer to held-out tasks.These gains, however, do not seem to constitute introspection consistently: improved self-modeling may not arise from privileged access to the model's internal decision process.We release our evaluation and training code on GitHub. 1
Authors: Siqi Zeng, Andre N. Assis, Rowan Wang
Institution: University of Illinois Urbana-Champaign
Language: English
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7204937544