AI RESEARCH

TimeSAF: Towards LLM-Guided Semantic Asynchronous Fusion for Time Series Forecasting

arXiv CS.AI

ArXi:2604.12648v1 Announce Type: cross Despite the recent success of large language models (LLMs) in time-series forecasting, most existing methods still adopt a Deep Synchronous Fusion strategy, where dense interactions between textual and temporal features are enforced at every layer of the network.