This research explores how Large Language Models can interpret time series data by inferring natural language descriptions of underlying events - a critical capability for data-driven decision making.

• First study examining LLMs' ability to translate numerical time patterns into explanatory narratives
• Demonstrates how AI can help identify what caused specific patterns in time-based data
• Works across multiple domains including finance and healthcare
• Creates a bridge between quantitative data and human-understandable explanations

Medical Impact: In healthcare settings, this technology could transform how clinicians interpret patient monitoring data, helping identify and describe critical events in patient conditions without requiring specialized technical expertise.

Inferring Event Descriptions from Time Series with Language Models