Research map / Large language models / Agents, code and math
A Principle-Driven Adaptive Policy for Group Cognitive Stimulation Dialogue for Elderly with Cognitive Impairment
Jiyue Jiang, Yanyu Chen, Pengan Chen, Kai Liu, Jingqi Zhou, Zheyong Zhu, He Hu, Fei Ma, Qi Tian, Chuan Wu
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 pre-training on simulated PGSS dialogues before real-data fine-tuning (ablation) | GCSD-3b · Real CST test set | BLEU-4: 26.51 -> 27.93 | Improved |
| add dynamic participant cognitive state modeling (soft prompt) (ablation) | GCSD-3b · Real CST test set | BLEU-4: 23.15 -> 27.93 | Improved |
| add cognitive stimulation-focused attention loss (ablation) | GCSD-3b · Real CST test set | BLEU-4: 24.98 -> 27.93 | Improved |
| add dynamic participant state modeling (ablation) | GCSD-3b · Real CST test set | Distinct-2: 60.13 -> 74.82 | Improved |
| add pre-training on simulated data (ablation) | GCSD-3b · Real CST test set | Distinct-2: 74.19 -> 74.82 | No clear effect |
| add dynamic participant state modeling (ablation) | GCSD-3b · Real CST test set | BERTScore: 79.48 -> 80.12 | No clear effect |
| GCSD (3B) vs strongest 5-shot baseline | Qwen2.5-3B · Real CST test set | BLEU-4: 24.36 -> 27.93 | Improved |
| GCSD vs GPT-4o in human A/B test | Qwen2.5-3B · Human A/B test | win rate: 50% win vs 39% loss | Improved |
| GCSD vs DeepSeek-671B in human A/B test | Qwen2.5-3B · Human A/B test | win rate: 43% win vs 40% loss | No clear effect |
Closest papers by meaning
- Zhongjing: Enhancing the Chinese Medical Capabilities of Large Language Model through Expert Feedback and Real-World Multi-Turn Dialogue (AAAI 2024 · similarity 0.83)
- TherapyGym: Evaluating and Aligning Clinical Fidelity and Safety in Therapy Chatbots (NeurIPS 2026 · similarity 0.83)
- Evaluating LLMs Across Multi-Cognitive Levels: From Medical Knowledge Mastery to Scenario-Based Problem Solving (ICML 2025 · similarity 0.82)
- SPASCA: Social Presence and Support with Conversational Agent for Persons Living with Dementia (AAAI 2025 · similarity 0.82)
- An LLM-based Simulation Framework for Embodied Conversational Agents in Psychological Counseling (AAAI 2026 · similarity 0.82)
- CareBot: A Pioneering Full-Process Open-Source Medical Language Model (AAAI 2025 · similarity 0.82)
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.
