发表机构
Deakin University; Applied Artificial Intelligence Initiative（迪肯大学; 应用人工智能倡议）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究文本条件时间序列预测问题，核心方法是将预测设为TSFM生成轨迹上的规划问题，以冻结TSFM为模拟器、LLM为策略和价值函数，实例化\rc{}框架，主要贡献是通过实验证明该框架能带来持续改进。
AI 中文摘要
文本条件时间序列预测可根据数值历史和自然语言上下文预测序列，这需要可靠的数值预测和上下文信息解释能力。时间序列基础模型（TSFMs）能提供强大的数值预测，大语言模型（LLMs）可对文本进行推理，但结合二者优势具有挑战性。为此将预测表述为TSFM生成轨迹上的规划问题，冻结的TSFM作为模拟器，LLM作为策略和价值函数。具体实例化为\rc{}（大语言模型作为预测规划器），这是一个免训练框架，通过蒙特卡洛树搜索弥合模态差距。在两个TSFM主干和四个LLMs上的实验表明，\rc{}在模型选择上带来持续改进，支持序列搜索作为文本条件预测的有效免训练方法。
英文摘要
Text-conditioned time-series forecasting predicts a series from both its numerical history and natural-language context, allowing forecasts to account for events and constraints that the past alone cannot reveal. This requires both reliable numerical forecasting and the ability to interpret contextual information. Time-series foundation models (TSFMs) provide strong numerical forecasts, while large language models (LLMs) can reason over text, but combining their strengths remains challenging because asking an LLM to generate or revise forecast values directly can distort the temporal structure captured by the TSFM. We instead formulate forecasting as a planning problem over TSFM-generated trajectories. The frozen TSFM acts as a simulator that proposes numerical continuations, while the LLM acts as a policy and value function that guides candidate selection and evaluates completed trajectories against the context. We instantiate this as \rc{} (\textbf{L}LM \textbf{A}s \textbf{F}orecasting \textbf{P}lanner), a training-free framework that bridges the modality gap without retraining either model, using Monte Carlo tree search (MCTS) over the forecast horizon with a \emph{Ranker} LLM as policy and a \emph{Judge} LLM as value function. Experiments on Context-is-Key and Time-MMD across two TSFM backbones (Chronos and TimesFM) and four LLMs show that \rc{} delivers consistent improvements across model choices, supporting sequential search as an effective training-free approach to text-conditioned forecasting.
