Shuhan Xiao

Meet OSF — a fully open benchmark and state-of-the-art family of sleep foundation models. 🌙 We study pre-training and scaling recipes that actually improve generalization in real-world settings. 🏥 🌐 Website: yang-ai-lab.github.io/osf/ 💻 Code: github.com/yang-ai-lab/OS… 🤗 Models: hf.co/yang-ai-lab/OS… 🧵👇

Can LLMs really reason over health time series? 📈 Introducing HEARTS ❤️— the first living benchmark built for health time-series reasoning. 🧪 20K+ test samples. 🧩 110 tasks. 🏥 12 health domains. 📡 20 signal modalities. 🌐Website: yang-ai-lab.github.io/HEARTS 🕵️Code: github.com/yang-ai-lab/HE… 🤗Dataset: huggingface.co/datasets/yang-… 🏆Leaderboard: yang-ai-lab.github.io/HEARTS/leaderb… 🧵👇

🌙 What if your sleep signals could speak? Introducing SleepLM — the first family of sleep-language foundation models for sleep physiology. SleepLM turns raw sleep physiology into something we can describe, query, and localize with language. 🧠Open-ended description 🔍Precise localization 🗣️Language-guided analysis 🌐Website: yang-ai-lab.github.io/SleepLM 🕵️Code: github.com/yang-ai-lab/Sl… 🤗Models: hf.co/yang-ai-lab/Sl… 🧵👇