AI Paper Atlas

Research map / Time series / Causal and generative time series

From Observations to States: Latent Time Series Forecasting

Jie Yang, Yifan Hu, Yuante Li, Kexin Zhang, Kaize Ding, Philip S. Yu

ICML 2026 · Time series › Causal and generative time series · approach: State-space and latent dynamics 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
Train backbone to forecast latent states (LatentTSF) instead of observationsiTransformer · Electricity (avg)MSE: 0.268 -> 0.194Improved
Train backbone to forecast latent states (LatentTSF)PatchTST · Electricity (avg)MSE: 0.389 -> 0.207Improved
Train backbone to forecast latent states (LatentTSF)TimeXer · Traffic (avg)MSE: 1.270 -> 0.636Improved
Train backbone to forecast latent states (LatentTSF)DLinear · Traffic (avg)MSE: 0.525 -> 0.552Hurt
Train backbone to forecast latent states (LatentTSF)iTransformer · ETTh1 (avg)MSE: 0.485 -> 0.441Improved
Train backbone to forecast latent states without Align loss versus vanilla DLinear (ablation)DLinear · ElectricityMSE: 8.8% MSE reductionImproved
Add alignment loss in latent space on top of latent prediction (ablation)LatentTSF DLinear · five datasetsMSE: ranking (a) > (b)Improved
Apply alignment loss in observation space instead (ablation)DLinear · five datasetsMSE: (c) approx (d)No clear effect
Add observation-space perceptual loss and fine-tune the autoencoder (ablation)LatentTSF iTransformer · ElectricityMSE: best 0.297 vs default 0.194Hurt
Replace pre-trained autoencoder with randomly initialised 2-layer MLPs (ablation)LatentTSF iTransformer · ElectricityMSE: best 0.351 vs 0.297 with pre-trained AEHurt
Increase Pred loss weight from 0 to moderate value (5), then 10 (ablation)LatentTSF · multiple datasetsMAE: substantial MAE reduction, marginal gain at 10Improved
Raise Align loss weight from 0 to 5 (ablation)LatentTSF · multiple datasetsMAE: sharp MAE drop, plateau beyond 5-10Improved

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