Due to privacy concerns, open dialogue datasets for mental health are primarily generated through human or AI synthesis methods. However, the inherent implicit nature of psychological processes, particularly those of clients, poses challenges to the authenticity and diversity of synthetic data. In this paper, we propose ECAs (short for Embodied Conversational Agents), a framework for embodied agent simulation based on Large Language Models (LLMs) that incorporates multiple psychological theoretical principles. Using simulation, we expand real counseling case data into a nuanced embodied cognitive memory space and generate dialogue data based on high-frequency counseling questions. We validated our framework using the D4 dataset. First, we created a public ECAs dataset through batch simulations based on D4. Licensed counselors evaluated our method, demonstrating that it significantly outperforms baselines in simulation authenticity and necessity. Additionally, two LLM-based automated evaluation methods were employed to confirm the higher quality of the generated dialogues compared to the baselines.

A significant number of De facto orphans in underdeveloped regions face potential mental health risks, while some novel interventions attempt to involve university student volunteers in providing companionship. However, there is little documented literature on current practices of such emotional support volunteer activities. We conducted a comprehensive investigation of multiple stakeholders involved in an online companionship program facilitated by university student volunteers, designed to provide remote emotional support to these children. Through field observations and interviews, we summarize current practices and identify benefits. We discovered that current remote volunteer initiatives face numerous challenges impeding companionship effectiveness. We summarized four potential technological requirements derived from these challenges. Our study provides the first documented account of remote volunteer companionship activities aimed at improving vulnerable children’s mental health. Our research can inspire future developments of technological solutions for improving emotional companionship in volunteer programs and vulnerable children’s wellbeing.

Online collaborative learning is increasingly important, yet children still face challenges communicating and working together virtually, limiting their engagement in long-term teamwork. To address this, we designed the Children’s Online Long-term Program (COLP), a 16-week online project-based learning program grounded in multiple learning theories. The program was implemented with 67 upper primary school students (Grades 3–6, ages 8–13) across five provinces in China. Results show that over one-third of participants sustained engagement in online teamwork. Interviews with children and their parents further revealed key communication channels, benefits, and challenges. Notably, parents played multiple roles in supporting their children’s collaboration, especially through modeling and guidance. This study contributes to the design of long-term online collaborative learning interventions for children within computer-supported collaborative learning (CSCL) communities.

Motion prediction is among the most fundamental tasks in autonomous driving. Traditional methods of motion forecasting primarily encode vector information of maps and historical trajectory data of traffic participants, lacking a comprehensive understanding of overall traffic semantics, which in turn affects the performance of prediction tasks. In this paper, we utilized Large Language Models (LLMs) to enhance the global traffic context understanding for motion prediction tasks. We first conducted systematic prompt engineering, visualizing complex traffic environments and historical trajectory information of traffic participants into image prompts— Transportation Context Map (TC-Map), accompanied by corresponding text prompts. Through this approach, we obtained rich traffic context information from the LLM. By integrating this information into the motion prediction model, we demonstrate that such context can enhance the accuracy of motion predictions. Furthermore, considering the cost associated with LLMs, we propose a cost-effective deployment strategy: enhancing the accuracy of motion prediction tasks at scale with 0.7% LLM-augmented datasets. Our research offers valuable insights into enhancing the understanding of traffic scenes of LLMs and the motion prediction performance of autonomous driving. The source code is available at https://github.com/AIR-DISCOVER/LLM-Augmented-MTR and https://aistudio.baidu.com/projectdetail/7809548.