On-Device Language Models for Privacy-Preserving Stress Prediction: A Multimodal Evaluation on Mobile Health

Abstract:Stress is a pervasive determinant of mental health and a key target for mobile health interventions. On-device language models (ODLMs) offer privacy-preserving inference without cloud dependency, yet their feasibility for health prediction under mobile resource constraints remains underexplored. We evaluate ODLMs for multi-modal stress prediction using zero-shot prompting, measuring predictive accuracy alongside latency and throughput. Our results show that objective sensor features marginally outperform subjective self-reports on average, and that lightweight sub-2B models achieve low latency with predictable resource usage. Our findings highlight both the promise and the practical constraints of ODLMs for mobile mental health.

6 pages, 2 figures. Received Honorable Mention at the 2026 Human-centered AI Research for Mental health, an Open Networking Symposium (HARMONY 2026) workshop, co-located with IEEE/ACM Conference on Connected Health: Applications, Systems, and Engineering Technologies (CHASE 2026)

Machine Learning (cs.LG); Human-Computer Interaction (cs.HC)

arXiv:2609.11961 [cs.LG]

(or arXiv:2609.11961v1 [cs.LG] for this version)