arXiv cs.AI· Sungjun Cho, Changho Shin, Suenggwan Jo, Xinya Yan, Shourjo Aditya Chaudhuri, Frederic Sala·· 14 小时前AI 评分31
LBS：LLM 与贝叶斯状态空间模型结合的多模态时间序列预测框架
LLM-Integrated Bayesian State Space Models for Multimodal Time-Series Forecasting
AI 导读
研究者提出 LLM 集成贝叶斯状态空间模型 LBS，将捕捉潜在状态时序动态的 SSM 主干与预训练 LLM 结合，由 LLM 编码文本输入做后验状态估计并解码与潜在轨迹一致的文本预测。该框架支持灵活的回顾与预测窗口，并提供有原则的不确定性量化。在 TextTimeCorpus 基准上，LBS 较此前 SOTA 提升 13.20%，同时为每次预测生成人类可读的摘要。
正文 · 原文
Abstract:Forecasting in the real world requires integrating structured time-series data with unstructured textual information, but existing methods are architecturally limited by fixed input/output horizons and are unable to model or quantify uncertainty. We address this challenge by introducing LLM-integrated Bayesian State space models (LBS), a novel probabilistic framework for multimodal temporal forecasting. At a high level, LBS consists of two components: (1) a state space model (SSM) backbone that captures the temporal dynamics of latent states from which both numerical and textual observations are generated and (2) a pretrained large language model (LLM) that is adapted to encode textual inputs for posterior state estimation and decode textual forecasts consistent with the latent trajectory. This design enables flexible lookback and forecast windows, principled uncertainty quantification, and improved temporal generalization thanks to the well-suited inductive bias of SSMs toward modeling dynamical systems. Experiments on the TextTimeCorpus benchmark demonstrate that LBS improves the previous state-of-the-art by 13.20% while providing human-readable summaries of each forecast. Our work is the first to unify LLMs and SSMs for joint numerical and textual prediction, offering a novel foundation for multimodal temporal reasoning.
| Comments: | 15 pages, 8 figures |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2510.20952 [cs.LG] |
| (or arXiv:2510.20952v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2510.20952 arXiv-issued DOI via DataCite |
Submission history
From: Sungjun Cho [view email]
[v1]
Thu, 23 Oct 2025 19:28:26 UTC (278 KB)
来源：arXiv cs.AI · arxiv.org