AI Paper Atlas

Research map / Time series / Foundation models for time series

ConFlux: Multivariate Time Series in Flux, One Unified Forecast in Confluence

Shiyu Wang, Juntong Ni, Ziyi Zhang, Baichuan Mo, Xinyue Zhong, Chengxin Wang, Yuchen Fang, Zhou Ye, Yang Xiang

ICML 2026 · Time series › Foundation models for time series · approach: Time series foundation models

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add variate sorting (Fiedler-vector ordering) before variate patching (ablation)ConFlux · ETTh1 zero-shotMSE: 0.397 -> 0.378Improved
add variate sorting (ablation)ConFlux · ETTm1 zero-shotMSE: 0.325 -> 0.314Improved
add variate sorting (ablation)ConFlux · Solar zero-shotMSE: 0.165 -> 0.162Improved
train with variate patching (ablation)ConFlux · Weather zero-shotMSE: 0.270 -> 0.229Improved
train with variate patching (ablation)ConFlux · ETTm1 zero-shotMSE: 0.366 -> 0.314Improved
use variate patching at inference (ablation)ConFlux · ETTh1 zero-shotMSE: 0.390 -> 0.378Improved
use variate patching at inference (ablation)ConFlux · Electricity zero-shotMSE: 0.165 -> 0.165No clear effect
use ConFlux instead of Chronos-2 for zero-shot long-term forecastingChronos-2 · long-term datasetsMSE: average 12.5% reduction; up to 41.1% on ElectricityImproved
use ConFlux instead of Chronos-2 on short-term trafficChronos-2 · PEMS03/04/07/08 zero-shotMSE: average 29.7% reduction; 39.3% on PEMS07Improved
in-domain fine-tuning vs multivariate Timer-XLTimer-XL · long-term fine-tunedMSE: average 30.7% reduction vs Timer-XLImproved
in-domain fine-tuning ConFlux vs ChronosChronos · Electricity fine-tunedMSE: 0.154 (Chronos) vs 0.159 (ConFlux)Hurt
scale up model size (Mini to Large) with more data (ablation)ConFlux · scalabilityMAEImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy