Researchers have developed a new benchmark to evaluate how well large language models (LLMs) can map multimodal human sensing data to psychological constructs. This benchmark, utilizing the Generalization of Longitudinal Behavior Modeling (GLOBEM) dataset, aligns passive sensing, ecological momentary assessment (EMA), and questionnaire evidence with 29 Brief Hierarchical Taxonomy of Psychopathology (B-HiTOP) items. The study found that while semantic abstraction aids in organizing self-report data, it can act as an information bottleneck for indirect behavioral sensing signals. AI
IMPACT This research could lead to more sophisticated LLM applications in mental health monitoring and analysis.
RANK_REASON The cluster contains a research paper detailing a new benchmark and methodology for evaluating LLMs. [lever_c_demoted from research: ic=1 ai=1.0]
- arXiv
- Brief Hierarchical Taxonomy of Psychopathology
- ecological momentary assessment
- Generalization of Longitudinal Behavior Modeling
- Hugging Face
- large language models
- multimodal human sensing
- questionnaire evidence
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
