Empowering VLMs for Few-Shot Multimodal Time Series Classification via Tailored Agentic Reasoning

In three linesMarsTSC, an agentic framework for VLMs, addresses few-shot multimodal time series classification. Three collaborative roles (Generator, Reflector, Modifier) iteratively refine a knowledge bank via reflective reasoning. Tested on 12 benchmarks across 6 VLM backbones: consistent gains, interpretable rationales.

Summary generated by Claude — human-verified