CastFSR: A Fast--Slow--Reflect Agentic Reasoning Framework for Context-Aware Time Series Forecasting
Quick Answer
CastFSR introduces a Fast--Slow--Reflect framework for context-aware time series forecasting, enhancing decision-making by integrating historical data with evolving contextual features.
Quick Take
It outperforms existing models through a two-stage strategy involving lightweight forecasters and reinforcement learning, demonstrating superior accuracy on public datasets.
Key Points
- CastFSR employs a Fast--Slow--Reflect workflow for improved forecasting accuracy.
- The framework adapts look-back windows based on contextual evidence.
- It supports training-free inference with existing .
- Extensive experiments show consistent outperformance against baseline models.
- Code for CastFSR is publicly available for further research.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Time series forecasting is fundamental to decision-making in complex systems, where future dynamics are influenced not only by historical observations but also by evolving contextual features. Recent advances in large language models (LLMs) have extended forecasting beyond numerical extrapolation toward context-aware reasoning. However, existing approaches often lack explicit mechanisms to identify relevant contexts, reason about their impacts, and validate forecasts against temporal and domain constraints. In this work, we propose CastFSR, an agentic framework that formulates context-aware forecasting as a Fast--Slow--Reflect workflow. In fast thinking, CastFSR profiles observations and selects lightweight forecasters to construct a data-driven forecast prior. In slow deliberation, it retrieves contextual evidence, adaptively determines informative look-back windows, and reasons about how contexts reshape future dynamics. In reflection, it iteratively refines forecasts to ensure temporal, contextual, and domain consistency. CastFSR supports both training-free inference with off-the-shelf LLMs and efficient deployment through a two-stage SFT and reinforcement learning strategy that transfers its orchestration capability to compact LLMs. Extensive experiments on public datasets demonstrate that CastFSR consistently outperforms representative baselines. Our code is available at this https URL.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2608.03031 [cs.AI] |
| (or arXiv:2608.03031v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2608.03031 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Xiaoyu Tao [view email]
[v1]
Tue, 4 Aug 2026 02:21:36 UTC (3,768 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.AI
See more →HOBA: Hierarchical On-Policy Bidding Agents for Adaptive Online Advertising
HOBA (Hierarchical On-policy Bidding Agents) is a novel hierarchical reinforcement learning framework that enhances online advertising bidding systems by improving adaptability and reducing hyperparameter tuning costs. It utilizes a for hyperparameter inference, a SARSA agent for expert model selection, and a dynamic expert pool for bid execution, achieving a +3.6% increase in target cost during large-scale deployment and outperforming state-of-the-art baselines on AuctionNet.