mlsanityData to 14 Oct 2026

MetaCaster: Meta-Harness-Optimized Agent for End-to-End Few-Shot Learning of Lightweight Time Series Forecasters

arXiv 2608.23473 ↗First posted 24 Aug 2026cs.LG cs.AIcited by 0

On the map: LLM agents and tool use → Financial trading and forecasting agents

Abstract (arXiv)

Time series forecasting (TSF) is evolving toward multimodal and agentic settings, yet using foundation models remains uneconomical in resource-constrained scenarios, where compact, specialized forecasters are more desirable. However, lightweight forecasters typically require substantial training data, limiting their use in domains with scarce, slowly accumulated, or privacy-sensitive time series. To address this dilemma, we investigate the challenging problem of few-shot learning for lightweight forecasters. We propose MetaCaster, a meta-harness-optimized multi-agent framework that uses agentic data generation to automatically train specialized lightweight forecasters from only a few examples and textual contexts. Our work highlights a new TSF paradigm in which agents act not as forecasters but as intermediary engineers that prepare efficient, task-specific forecasters for deployment. Experiments on 18 datasets, 23 state-of-the-art lightweight forecasters, and 14 baselines demonstrate that MetaCaster achieves both data efficiency and computational efficiency while maintaining high-quality TSF performance.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Financial trading and forecasting agents, one of 16 directions in LLM agents and tool use. Builds on 43 of its 67 references that are in the map. First posted on arXiv 24 Aug 2026.

Builds on

43 of 67 references.

Closest ancestors · 3

Same direction, newest first

  1. 2026Nexus : An Agentic Framework for Time Series Forecasting14 May
  2. 2026TimeART: Towards Agentic Time Series Reasoning via Tool-Augmentation20 Jan
  3. 2025FLAIRR-TS -- Forecasting LLM-Agents with Iterative Refinement and Retrieval for Time Series24 Aug

Other references · 40

  1. 2020Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecastingcited by 2,169
  2. 2022A Time Series is Worth 64 Words: Long-term Forecasting with Transformerscited by 2,039
  3. 2024Chronos: Learning the Language of Time Seriescited by 759
  4. 2023One Fits All:Power General Time Series Analysis by Pretrained LMcited by 632
  5. 2023Large Language Models Are Zero-Shot Time Series Forecasterscited by 500
  6. 2021TS2Vec: Towards Universal Representation of Time Seriescited by 471
  7. 2018T-GCN: A Temporal Graph ConvolutionalNetwork for Traffic Predictioncited by 401
  8. 2026SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Taskscited by 315
  9. 2026Meta-Harness: End-to-End Optimization of Model Harnessescited by 309
  10. 2025Small Language Models are the Future of Agentic AIcited by 270
  11. 2021Monash Time Series Forecasting Archivecited by 267
  12. 2020Time Series Data Augmentation for Deep Learning: A Surveycited by 226
28 more
  1. 2024GIFT-Eval: A Benchmark For General Time Series Forecasting Model Evaluationcited by 212
  2. 2023FITS: Modeling Time Series with 10k Parameterscited by 195
  3. 2024Diffusion-TS: Interpretable Diffusion for General Time Series Generationcited by 184
  4. 2025Sundial: A Family of Highly Capable Time Series Foundation Modelscited by 180
  5. 2022Less Is More: Fast Multivariate Time Series Forecasting with Light Sampling-oriented MLP Structurescited by 160
  6. 2021TimeVAE: A Variational Auto-Encoder for Multivariate Time Series Generationcited by 142
  7. 2026Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnessescited by 142
  8. 2020An Empirical Survey of Data Augmentation for Time Series Classification with Neural Networkscited by 128
  9. 2024CALF: Aligning LLMs for Time Series Forecasting via Cross-modal Fine-Tuningcited by 108
  10. 2024SparseTSF: Modeling Long-term Time Series Forecasting with 1k Parameterscited by 107
  11. 2024CycleNet: Enhancing Time Series Forecasting through Modeling Periodic Patternscited by 106
  12. 2024ChatTime: A Unified Multimodal Time Series Foundation Model Bridging Numerical and Textual Datacited by 100
  13. 2024Language Models Still Struggle to Zero-shot Reason about Time Seriescited by 88
  14. 2024VisionTS: Visual Masked Autoencoders Are Free-Lunch Zero-Shot Time Series Forecasterscited by 88
  15. 2025Time-MQA: Time Series Multi-Task Question Answering with Context Enhancementcited by 82
  16. 2021Time Series Prediction using Deep Learning Methods in Healthcarecited by 74
  17. 2023Time Series Contrastive Learning with Information-Aware Augmentationscited by 52
  18. 2024Utilizing Image Transforms and Diffusion Models for Generative Modeling of Short and Long Time Seriescited by 43
  19. 2024xPatch: Dual-Stream Time Series Forecasting with Exponential Seasonal-Trend Decompositioncited by 36
  20. 2025Multi-modal Time Series Analysis: A Tutorial and Surveycited by 24
  21. 2025CrossLinear: Plug-and-Play Cross-Correlation Embedding for Time Series Forecasting with Exogenous Variablescited by 22
  22. 2025TimeDP: Learning to Generate Multi-Domain Time Series with Domain Promptscited by 21
  23. 2025T2S: High-resolution Time Series Generation with Text-to-Series Diffusion Modelscited by 20
  24. 2025TimeEmb: A Lightweight Static-Dynamic Disentanglement Framework for Time Series Forecastingcited by 12
  25. 2025TS-Reasoner: Aligning Time Series Foundation Models with LLM Reasoningcited by 12
  26. 2025AlphaCast: A Human Wisdom-LLM Intelligence Co-Reasoning Framework for Interactive Time Series Forecastingcited by 12
  27. 2025CMoS: Rethinking Time Series Prediction Through the Lens of Chunk-wise Spatial Correlationscited by 9
  28. 2025SVTime: Small Time Series Forecasting Models Informed by "Physics" of Large Vision Model Forecasterscited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025TimeSeriesScientist: A General-Purpose AI Agent for Time Series Analysis
  2. 2026Cast-R1: Learning Tool-Augmented Sequential Decision Policies for Time Series Forecasting
  3. 2026CastFSR: A Fast--Slow--Reflect Agentic Reasoning Framework for Context-Aware Time Series Forecasting
  4. 2026Position: Beyond Model-Centric Prediction -- Agentic Time Series Forecasting
  5. 2026A Human-in-the-Loop Autonomous Agent for Industry Time Series Forecasting
  6. 2025TimeCopilot
  7. 2026MACTS-EM: Multi-Agent Collaborative Time Series Forecasting with Emergent Memory
  8. 2026When Tomorrow Becomes Today: Self-Evolving Policies for Agentic Time-Series Forecasting