Can LLM Embeddings Improve Time Series Forecasting? A Practical Feature Engineering Approach

Using large language models (LLMs) — or their outputs, for that matter — for all kinds of machine learning-driven tasks, including predictive ones that were already being solved long before...