ChronoSteer: Bridging Large Language Model and Time Series Foundation Model via Synthetic Data

2025/05/15 by Chengsen Wang, Qi Qi, Wang, Chengsen +9 · 1 citation
Computer Science · #FOS: Computer and information sciences #Machine Learning (cs.LG) #Topic Modeling

paper · pdf · doi:10.48550/arxiv.2505.10083

openalex publication_date 2025/05/15 · openalex created_date 2025/10/15 · openalex updated_date 2026/07/28

Abstract

Conventional forecasting methods rely on unimodal time series data, limiting their ability to exploit rich textual information. Recently, large language models (LLMs) and time series foundation models (TSFMs) have demonstrated powerful capability in textual reasoning and temporal modeling, respectively. Integrating the strengths of both to construct a multimodal model that concurrently leverages both temporal and textual information for future inference has emerged as a critical research challenge. To address the scarcity of event-series paired data, we propose a decoupled framework: an LLM is employed to transform textual events into revision instructions, which are then used to steer the output of TSFM. To implement this framework, we introduce ChronoSteer, a multimodal TSFM that can be steered through textual revision instructions, effectively bridging LLM and TSFM. Moreover, to mitigate the shortage of cross-modal instruction-series paired data, we devise a two-stage training strategy based on synthetic data. In addition, we also construct a high-quality multimodal time series forecasting benchmark to address the information leakage concerns during evaluation. After integrating with an LLM, ChronoSteer, which is trained exclusively on synthetic data, achieves a 25.7% improvement in prediction accuracy compared to the unimodal backbone and a 22.5% gain over the previous state-of-the-art multimodal method.

Citations

• DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning
• ChatTime: A Unified Multimodal Time Series Foundation Model Bridging Numerical and Textual Data
• A Picture is Worth A Thousand Numbers: Enabling LLMs Reason about Time Series via Visualization
• Context is Key: A Benchmark for Forecasting with Essential Textual Information
• TimeSeriesExam: A time series understanding exam
• Intervention-Aware Forecasting: Breaking Historical Limits from a System Perspective
• Evaluating Large Language Models on Time Series Feature Understanding: A Comprehensive Taxonomy and Benchmark
• Language Models Still Struggle to Zero-shot Reason about Time Series
• S²IP-LLM: Semantic Space Informed Prompt Learning with LLM for Time Series Forecasting
• LSTPrompt: Large Language Models as Zero-Shot Time Series Forecasters by Long-Short-Term Prompting
• MOMENT: A Family of Open Time-series Foundation Models
• M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillation
• Unified Training of Universal Time Series Forecasting Transformers
• Tiny Time Mixers (TTMs): Fast Pre-trained Models for Enhanced Zero/Few-Shot Forecasting of Multivariate Time Series
• A decoder-only foundation model for time-series forecasting
• Large Language Models Are Zero-Shot Time Series Forecasters
• iTransformer: Inverted Transformers Are Effective for Time Series Forecasting
• Visual Instruction Tuning
• GPT-4 Technical Report
• One Fits All:Power General Time Series Analysis by Pretrained LM
• A Time Series is Worth 64 Words: Long-term Forecasting with Transformers
• TimesNet: Temporal 2D-Variation Modeling for General Time Series Analysis
• Are Transformers Effective for Time Series Forecasting?
• Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting
• Learning Transferable Visual Models From Natural Language Supervision
• Ensemble learning for electricity consumption forecasting in office buildings
• Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting
• Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting
• Forecasting Intra-Hour Imbalances in Electric Power Systems
• DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networks
• Long Short-Term Memory
• Towards Explainable Traffic Flow Prediction with Large Language Models
• Time-MMD: Multi-Domain Multimodal Dataset for Time Series Analysis

Cited by

• TS-Reasoner: Aligning Time Series Foundation Models with LLM Reasoning

Related

• MAP4TS: A Multi-Aspect Prompting Framework for Time-Series Forecasting with Large Language Models
• From Text to Time? Rethinking the Effectiveness of the Large Language Model for Time Series Forecasting
• DP-GPT4MTS: Dual-Prompt Large Language Model for Textual-Numerical Time Series Forecasting
• Large Language Models Are Zero-Shot Time Series Forecasters
• CC-Time: Cross-Model and Cross-Modality Time Series Forecasting