In‑context Time Series Predictor
PDF OpenReview arXiv Code Poster Slides
Reformulates TSF as in‑context learning by constructing tokens of (lookback, future) task pairs, enabling Transformers to adapt predictors from context without parameter updates.
Applies core ICL principles to TSF, turning forecasting tasks into tokens and leveraging context to induce suitable predictors with efficient, non‑LLM‑scale models.