arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2601.19151cs.AIcs.MA
收藏

TS-Debate:多模态协作辩论用于零样本时间序列推理

Multimodal Collaborative Debate for Zero-Shot Time Series Reasoning

KAIST Seoul(韩国科学技术院)

Patara Trirat, Jin Myung Kwak, Jay Heo, Heejun Lee, Sung Ju Hwang

AI总结:

TS-Debate通过多模态协作辩论框架,在零样本时间序列推理中实现显著性能提升,通过专门代理和结构化协议提升模态保真度和数值精度。

摘要:

近年来,大语言模型(LLMs)与时间序列(TS)分析的交叉进展揭示了潜力与脆弱性。尽管LLMs可以在精心设计的上下文中推理时间结构,但它们在数值精度、模态干扰和原理性跨模态整合方面常常面临挑战。我们提出了TS-Debate,一种专门化、协作的多代理辩论框架,用于零样本时间序列推理。TS-Debate为文本上下文、视觉模式和数值信号分配专门的专家代理,并在显式领域知识提取后协调其交互,通过结构化的辩论协议。评审代理使用验证-冲突-校准机制评估代理主张,支持轻量级代码执行和数值查找以进行程序验证。该架构保持了模态保真度,暴露了冲突证据,并在不进行任务特定微调的情况下缓解了数值幻觉。在20个跨越三个公开基准的任务中,TS-Debate在强基线模型上实现了一致且显著的性能提升,包括所有代理观察所有输入的标准多模态辩论。

英文摘要:

Large language models (LLMs) are increasingly used as natural-language interfaces to structured data, yet they remain brittle when reasoning over time series. Visual patterns can be misleading, numerical claims can be hallucinated, and textual context can override evidence from the signal. We study zero-shot time-series reasoning as a multimodal evidence arbitration problem for LLM agents. We propose TS-Debate, an inference-time multi-agent protocol that requires no task-specific fine-tuning. TS-Debate first elicits relevant domain knowledge, then assigns modality-specialized agents to textual context, visual patterns, and numerical signals, and coordinates their interaction through a verification-conflict-calibration procedure. Reviewer agents check decision-critical claims with lightweight code execution and numerical lookup, resolve cross-modal disagreement, and calibrate the final answer. Unlike generic multi-agent debate or unconstrained tool use, TS-Debate specifies how evidence is exposed, which claims are checkable, and how verification outcomes shape synthesis. Across 20 tasks from three public benchmarks, TS-Debate improves classification and question answering performance over strong baselines, while revealing that debate is most useful for global-structure and cross-view reasoning rather than local value reconstruction.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑