Large Language Models (LLMs) have demonstrated effectiveness as zero-shot time series (TS) forecasters. The key challenge lies in tokenizing TS data into textual representations that align with LLMs' pre-trained knowledge. While existing work often relies on fine-tuning specialized modules to bridge this gap, a distinct, yet challenging, paradigm aims to leverage truly off-the-shelf LLMs without any fine-tuning whatsoever, relying solely on strategic tokenization of numerical sequences. The performance of these fully frozen models is acutely sensitive to the textual representation of the input data, as their parameters cannot adapt to distribution shifts. In this paper, we introduce a simple yet highly effective strategy to overcome this brittleness: injecting noise into the raw time series before tokenization. This non-invasive intervention acts as a form of inference-time augmentation, compelling the frozen LLM to extrapolate based on robust underlying temporal patterns rather than superficial numerical artifacts. We theoretically analyze this phenomenon and empirically validate its effectiveness across diverse benchmarks. Notably, to fully eliminate potential biases from data contamination during LLM pre-training, we introduce two novel TS datasets that fall outside all utilized LLMs' pre-training scopes, and consistently observe improved performance. This study provides a further step in directly leveraging off-the-shelf LLMs for time series forecasting.
InA-Probe: Instruction-Aware Active Probing for Time Series Forecasting with LLMs
Nanyang Technological University(南洋理工大学);Khalifa University(哈利法大学);Nanjing University of Aeronautics and Astronautics(南京航空航天大学);Singapore University of Technology and Design(新加坡科技设计大学)
UC Berkeley(加州大学伯克利分校);UC Santa Barbara(加州大学圣巴巴拉分校);Duke University(杜克大学);National University of Singapore(新加坡国立大学);King Abdulaziz City for Science and Technology(国王阿卜杜勒阿齐兹城市科学技术学院);University of Washington(华盛顿大学)
TAMMs: Change Understanding and Forecasting in Satellite Image Time Series with Temporal-Aware Multimodal Models
School of Computer Science & Technology, Beijing Institute of Technology(计算机科学与技术学院,北京理工大学);School of Computing, National University of Singapore(计算学院,新加坡国立大学)
Integrating Time Series into LLMs via Multi-layer Steerable Embedding Fusion for Enhanced Forecasting
School of Software Engineering Sun Yat-Sen University(软件工程学院中山大学);Institute of Data Science(数据科学研究院);School of Computing(计算机学院);National University of Singapore(新加坡国立大学)
Re(Visiting) Time Series Foundation Models in Finance
Alliance Manchester Business School University of Manchester(曼彻斯特大学阿利安斯商学院);Department of Mathematics University College London (UCL)(伦敦大学学院数学系);School of Economics Shanghai University(上海大学经济学院)
iTFKAN: Interpretable Time Series Forecasting with Kolmogorov-Arnold Network
Hong Kong Polytechnic University(香港理工大学);Northwestern Polytechnical University(西北工业大学);Sun Yat-sen University(中山大学);Hong Kong University of Science and Technology (Guangzhou)(香港科技大学(广州))
Understanding Token-level Topological Structures in Transformer-based Time Series Forecasting
University of Chinese Academy of Sciences(中国科学院大学);National Key Laboratory of Space Integrated Information System, Institute of Software, Chinese Academy of Sciences(空间信息集成系统国家重点实验室,软件研究所,中国科学院);Wangxuan Institute of Computer Technology, Peking University(王轩计算机技术研究所,北京大学);Hong Kong University of Science and Technology(香港科技大学)
Backdoor Samples Detection Based on Perturbation Discrepancy Consistency in Pre-trained Language Models
Key Laboratory of Aerospace Information Security and Trusted Computing, Ministry of Education, School of Cyber Science and Engineering, Wuhan University(航天信息安全部门、教育部、网络安全科学与工程学院、武汉大学)
TS-LIF: A Temporal Segment Spiking Neuron Network for Time Series Forecasting
Nanyang Technological University(南洋理工大学);University of Chinese Academy of Sciences(中国科学院大学);Tencent AI Lab(腾讯AI实验室);Webank-NTU Joint Research Institute on Fintech(微众银行-南洋理工大学金融科技联合研究院)
Institute for AI Industry Research (AIR), Tsinghua University(清华大学人工智能产业研究院(AIR));Shanghai Artificial Intelligence Laboratory(上海人工智能实验室);University of Notre Dame(圣母大学);Microsoft Research(微软研究院);Department of Automation, Tsinghua University(清华大学自动化系)
Robustness Test for AI Forecasting of Hurricane Florence Using FourCastNetv2 and Random Perturbations of the Initial Condition
Department of Computer Science, San Diego State University, San Diego, California, United States(计算机科学系,圣地亚哥州立大学,圣地亚哥,加利福尼亚州,美国);Department of Mathematics, San Diego State University, San Diego, California, United States(数学系,圣地亚哥州立大学,圣地亚哥,加利福尼亚州,美国)