CTRL：基于控制的时间序列预测与LLM引导的残差学习
CTRL: Control-Based Time Series Forecasting with LLM-Guided Residual Learning
- Yonsei University（延世大学）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
CTRL框架通过冻结骨干网络生成基础预测，利用LLM智能体分析残差并输出控制信号进行修正，实现语义推理与定量预测解耦，提升非平稳时间序列预测的鲁棒性。
AI中文摘要：
时间序列预测支撑着跨多个领域的关键决策。虽然大型语言模型（LLMs）提供了有前景的推理能力，但现有的基于LLM的时间序列预测方法要么将其简化为绕过其优势的数值预测器，要么允许直接生成预测，这在非平稳环境中会破坏预测的稳定性。我们引入了CTRL，一个将语义推理与定量预测解耦的框架。一个冻结的骨干网络生成基础预测，而专门的LLM智能体作为控制器，通过分解的趋势、季节性和不规则成分分析骨干网络的预测误差，将推理锚定在可解释的时间结构上。每个智能体输出紧凑的控制信号，由轻量级残差解码器转换为预测修正。CTRL包含无标签的测试时自适应，仅从输入统计中检测分布偏移，并通过缓存仅需3-24次LLM调用即可重新调整控制信号。CTRL明确设计用于提高非平稳时间动态和分布偏移下的鲁棒性，同时在高度平稳的时间序列上保持竞争力，此时自适应修正提供的额外收益有限。
英文摘要：
Time series forecasting underpins critical decision-making across diverse domains. While large language models (LLMs) offer promising reasoning capabilities, existing LLM-based time series forecasting approaches either reduce them to numerical predictors that bypass their strengths, or allow direct forecast generation that destabilizes predictions in non-stationary settings. We introduce CTRL, a framework that decouples semantic reasoning from quantitative prediction. A frozen backbone generates base forecasts, while specialized LLM agents function as controllers that analyze backbone prediction errors through decomposed trend, seasonal, and irregular components, grounding reasoning in interpretable temporal structure. Each agent outputs compact control signals that a lightweight residual decoder translates into forecast corrections. CTRL incorporates label-free test-time adaptation that detects distribution shift from input statistics alone and readapts control signals with only 3-24 LLM calls via caching. CTRL is explicitly designed to improve robustness under non-stationary temporal dynamics and distribution shift, while remaining competitive on highly stationary time series where adaptive correction provides limited additional benefit.
