SCALER predicts a coarse future shape with a lightweight Transformer, then lets a frozen LLM refine it in fixed steps, improving forecast accuracy and cutting inference cost versus test-time scaling baselines.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Efficient Test-Time Scaling for LLM-based Time Series Forecasting
SCALER predicts a coarse future shape with a lightweight Transformer, then lets a frozen LLM refine it in fixed steps, improving forecast accuracy and cutting inference cost versus test-time scaling baselines.