GlyRAG combines LLM-generated context, patch-based CGM representations, and retrieval of similar historical episodes to improve long-horizon blood glucose forecasting.
This study investigates the acute effects of different doses of cannabis on multiple stress-related systems, including subjective stress, cortisol, electrodermal activity, heart rate, and heart rate variability, using a randomized, double-blind, placebo-controlled design.
We propose a parameter-efficient continual learning method for wearable human activity recognition. By freezing pretrained features and learning lightweight channel-wise gates, the model adapts to new users while reducing catastrophic forgetting. On PAMAP2, it improved final accuracy from 56.7% to 77.7% and reduced forgetting from 39.7% to 16.2%, while training under 2% of the model’s parameters and requiring no replay buffer.