TS-Debate: Multimodal Collaborative Debate for Zero-Shot Time Series Reasoning

TL;DR

TS-Debate introduces a multi-agent debate framework that enhances zero-shot time series reasoning by preserving modality fidelity and mitigating hallucinations, leading to improved performance across multiple benchmarks.

Contribution

The paper presents a novel multi-agent debate system with dedicated modality experts and a structured protocol, improving zero-shot time series reasoning without task-specific fine-tuning.

Findings

Achieves significant performance gains over baselines on 20 tasks.

Effectively preserves modality fidelity and reduces numeric hallucinations.

Demonstrates robustness across diverse multimodal benchmarks.

Abstract

Recent progress at the intersection of large language models (LLMs) and time series (TS) analysis has revealed both promise and fragility. While LLMs can reason over temporal structure given carefully engineered context, they often struggle with numeric fidelity, modality interference, and principled cross-modal integration. We present TS-Debate, a modality-specialized, collaborative multi-agent debate framework for zero-shot time series reasoning. TS-Debate assigns dedicated expert agents to textual context, visual patterns, and numerical signals, preceded by explicit domain knowledge elicitation, and coordinates their interaction via a structured debate protocol. Reviewer agents evaluate agent claims using a verification-conflict-calibration mechanism, supported by lightweight code execution and numerical lookup for programmatic verification. This architecture preserves modality fidelity, exposes conflicting evidence, and mitigates numeric hallucinations without task-specific fine-tuning. Across 20 tasks spanning three public benchmarks, TS-Debate achieves consistent and significant performance improvements over strong baselines, including standard multimodal debate in which all agents observe all inputs.