AI Paper Atlas

Research map / Time series / Foundation models for time series

TimeCAP: A Channel-Aware Pre-Training Framework for Multivariate Time Series Forecasting

Chuanru Ren, Yao Lu, Tianjin Huang, Haowen Zheng, Hengde Zhu, Yunyin Li, Hengxiao Li, Lu Liu

AAAI 2026 · Time series › Foundation models for time series · approach: LLM-based time series forecasting · 0 citations (OpenAlex)

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 the autoregressive decoding head (ablation)TimeCAP · Electricity (full-shot)MSE: 0.186 -> 0.171Improved
add the autoregressive decoding head (ablation)TimeCAP · Exchange (full-shot)MSE: 0.347 -> 0.326Improved
add the one-shot decoding head (ablation)TimeCAP · Electricity (full-shot)MSE: 0.497 -> 0.171Improved
add the one-shot decoding head (ablation)TimeCAP · ETT-avg (full-shot)MSE: 0.399 -> 0.361Improved
add adaptive meta-routers and channel-aware cross-attention (ablation)TimeCAP · Exchange (full-shot)MSE: 0.349 -> 0.326Improved
add adaptive meta-routers and channel-aware cross-attention (ablation)TimeCAP · Electricity (full-shot)MSE: 0.181 -> 0.171Improved
use sigmoid-based fusion instead of direct weighted averaging (ablation)TimeCAP · Electricity (full-shot)MSE: 0.252 -> 0.171Improved
use sigmoid-based fusion instead of direct weighted averaging (ablation)TimeCAP · Weather (full-shot)MSE: 0.251 -> 0.250No clear effect
add self-supervised multi-domain pre-trainingTimeCAP · Electricity (full-shot)MSE: 0.244 -> 0.171Improved
add self-supervised pre-training before zero-shot forecastingTimeCAP · ETTh-avg (zero-shot)MSE: 0.631 -> 0.565Improved
add pre-training on 10% few-shot ETTh forecasting versus Timer-XLTimer-XL · ETTh-avg (few-shot 10%)MSE: 0.566 -> 0.474Improved
use TimeCAP instead of Timer-XL for efficiencyTimer-XL · ETTh2 zero-shot, horizon 720inference time (s): 9.97 -> 3.63Improved

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