Research map / Large language models / AI and society
ChatCLIDS: Simulating Persuasive AI Dialogues to Promote Closed-Loop Insulin Adoption in Type 1 Diabetes Care
Zonghai Yao, Talha Chafekar, Junda Wang, Shuo Han, Feiyun Ouyang, Junhui Qian, Lingxi Li, Hong Yu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add Chain of Strategy prompting (ablation) | GPT-4o nurse agent · ChatCLIDS Single-Visit (avg) | normalized persuasion rating: 0.214 -> 0.490 | Improved |
| add Chain of Strategy prompting (ablation) | Phi4 nurse agent · ChatCLIDS Single-Visit (avg) | normalized persuasion rating: -0.072 -> 0.193 | Improved |
| add Chain of Strategy prompting (ablation) | Deepseek-R1 nurse agent · ChatCLIDS Single-Visit (avg) | normalized persuasion rating: 0.372 -> 0.440 | Improved |
| add Chain of Strategy prompting (ablation) | Qwen2.5-0.5B nurse agent · ChatCLIDS Single-Visit (avg) | normalized persuasion rating: -0.291 -> -0.143 | Improved |
| add Chain of Strategy prompting (ablation) | all nurse models · ChatCLIDS Single-Visit, Hard patients | normalized persuasion rating: GPT-4o Hard -0.229 -> -0.075 | No clear effect |
| add Chain of Strategy prompting (ablation) | Deepseek-R1 nurse agent · ChatCLIDS Single-Visit, Easy patients | normalized persuasion rating: 0.940 -> 0.920 | Hurt |
| increase patient difficulty from Easy to Hard | nurse agents · ChatCLIDS Single-Visit | normalized persuasion rating: Deepseek-R1 0.940 -> -0.153 (Direct) | Hurt |
| add reflection and reasoning (o4-mini) versus non-reflective model | nurse agent · ChatCLIDS Multi-Visit | persuasion AUC: GPT-4o 17.84 -> o4-mini 46.50 | Improved |
| add adversarial Social Resistance agent | o4-mini · ChatCLIDS Multi-Visit vs Social Resistance | persuasion AUC: 46.50 -> 30.83 | Hurt |
| add adversarial Social Resistance agent | Deepseek-R1 · ChatCLIDS Multi-Visit vs Social Resistance | persuasion AUC: 36.66 -> 29.52 | Hurt |
| add adversarial Social Resistance agent | GPT-4.1 · ChatCLIDS Multi-Visit vs Social Resistance | persuasion AUC: 16.42 -> 25.07 | Mixed |
| scale model size within Qwen2.5 (ablation) | Qwen2.5 nurse agent · ChatCLIDS Single-Visit (Direct avg) | normalized persuasion rating: -0.291 (0.5B) -> 0.271 (32B) | Improved |
Closest papers by meaning
- TherapyGym: Evaluating and Aligning Clinical Fidelity and Safety in Therapy Chatbots (NeurIPS 2026 · similarity 0.85)
- CARES: A Comprehensive Benchmark of Trustworthiness in Medical Vision Language Models (NeurIPS 2024 · similarity 0.84)
- Measuring And Improving Persuasiveness Of Large Language Models (ICLR 2025 · similarity 0.84)
- An LLM-based Simulation Framework for Embodied Conversational Agents in Psychological Counseling (AAAI 2026 · similarity 0.83)
- SocialSim: Towards Socialized Simulation of Emotional Support Conversation (AAAI 2025 · similarity 0.83)
- CareBot: A Pioneering Full-Process Open-Source Medical Language Model (AAAI 2025 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
