Beyond Prompt: Fine-grained Simulation of Cognitively Impaired Standardized Patients via Stochastic Steering

2026-04-13

AI Summary

Existing approaches rely on discrete prompt engineering, which struggles to capture the fine-grained heterogeneity of cognitive impairment manifestations across diverse domains and severity levels. This work proposes a novel method that extracts guidance vectors from instruction–response contrastive pairs and integrates them with a Stochastic Token Modulation (STM) mechanism to enable controllable and nuanced simulation of cognitive impairments in language model outputs. By innovatively incorporating domain-specific guidance vectors and stochastic intervention probabilities, the approach significantly enhances both clinical fidelity and precise control over impairment severity. Experimental results demonstrate that the proposed method substantially outperforms existing baselines in terms of clinical realism and controllability.

Technology Category

Cognitive Modeling & Cognitive Systems: Simulating Human BehaviorMachine Learning: Large Multimodal Models (LMMs)Natural Language Processing: Prompt Engineering / Prompting

Application Category

Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactionsSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for rankingUser Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

Simulating Standardized Patients with cognitive impairment offers a scalable and ethical solution for clinical training. However, existing methods rely on discrete prompt engineering and fail to capture the heterogeneity of deficits across varying domains and severity levels. To address this limitation, we propose StsPatient for the fine-grained simulation of cognitively impaired patients. We innovatively capture domain-specific features by extracting steering vectors from contrastive pairs of instructions and responses. Furthermore, we introduce a Stochastic Token Modulation (STM) mechanism to regulate the intervention probability. STM enables precise control over impairment severity while mitigating the instability of conventional vector methods. Comprehensive experiments demonstrate that StsPatient significantly outperforms baselines in both clinical authenticity and severity controllability.

Problem

• cognitive impairment
• standardized patients
• heterogeneity
• severity levels
• clinical simulation

Innovation

• Stochastic Token Modulation
• steering vectors
• fine-grained simulation
• cognitively impaired standardized patients
• severity controllability