Search
ByteDance and Tsinghua University have collaboratively introduced ChatTS, an innovative temporal multi-modal large model. Boasting advanced capabilities in multivariate temporal question-answering and reasoning, ChatTS transcends the constraints of conventional models in processing temporal data. Leveraging a "purely synthetic-driven" methodology, ChatTS excels at generating diverse temporal data, analyzing complex multivariate temporal patterns, detecting fluctuation trends, and extracting anomalies. This versatility makes it ideal for a range of applications, including AIOps and finance, underscoring its vast potential. The groundbreaking research findings have been featured in VLDB 2025, further emphasizing its significant academic contributions and promising application prospects.
-
C114 Communication Network -
Communication Home
