כתבה
arXiv cs.AI ·
HealthSLM-Bench: בדיקת דגמי שפה קטנים למעקב רפואי בגדלים ניידים
HealthSLM-Bench: Benchmarking Small Language Models for Mobile and Wearable Healthcare Monitoring
בדיקת דגמי שפה קטנים למעקב רפואי בגדלים ניידים. נמצא שדגמי שפה קטנים יכולים להשיג תוצאות דומות לדגמי שפה גדולים, עם זאת, יש עדיין דרישות רבות.
תקציר מקורי באנגליתarXiv:2509.07260v5 Announce Type: replace Abstract: Mobile and wearable healthcare monitoring play a vital role in facilitating timely interventions, managing chronic health conditions, and ultimately improving individuals' quality of life. Previous studies on large language models (LLMs) have highlighted their impressive generalization abilities and effectiveness in healthcare prediction tasks. However, most LLM-based healthcare solutions are cloud-based, which raises significant privacy concerns and results in increased memory usage and latency. To address these challenges, there is growing interest in compact models, Small Language Models (SLMs), which are lightweight and designed to run locally and efficiently on mobile and wearable devices. Nevertheless, how well these models perform
קרא במקור המקורי
arxiv.org
פתח כתבה מקורית