AI Paper Atlas

Research map / Time series / Long-term forecasting

Local Geometry Attention for Time Series Forecasting under Realistic Corruptions

Dongbin Kim, Youngjoo Park, Woojin Jeong, Jaewook Lee

ICLR 2026 · Time series › Long-term forecasting · approach: Attention for time series forecasting

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
replace scaled dot-product attention with LGAPatchTST · ETTm1 combined severity 5MSE: 0.839 -> 0.734Improved
replace scaled dot-product attention with LGAPatchTST · ETTm1 combined severity 3MSE: 0.614 -> 0.519Improved
replace scaled dot-product attention with LGAPatchTST · Weather combined severity 5MSE: 0.491 -> 0.454Improved
replace scaled dot-product attention with LGAPatchTST · Electricity combined severity 5MSE: 0.290 -> 0.284Improved
replace scaled dot-product attention with LGAPatchTST · ETTm1 clean (severity 0)MSE: 0.352 -> 0.351No clear effect
replace TimeMixer with PatchLGAPatchLGA · Weather combined severity 5MSE: 0.576 -> 0.454Improved
replace standard attention with MoM robust attentionPatchTST · ETTm1 combined severity 5MSE: 0.839 -> 1.016Hurt
replace standard attention with Elliptical attentionPatchTST · ETTm1 combined severity 5MSE: 0.839 -> 0.880Hurt
replace SDP with LGA in CATS (cross-attention)CATS · ETTm1 combined severity 5MSE: 1.102 -> 1.037Improved
replace SDP with LGA in iTransformer (channel attention)iTransformer · ETTm1 combined severity 5MSE: 1.309 -> 1.266Improved
replace SDP with LGA under level shift onlyPatchTST · ETTm1 level shift severity 5MSE: 0.848 -> 0.718Improved
replace SDP with LGA under spike corruption onlyPatchTST · ETTm1 spike severity 5MSE: 0.374 -> 0.372No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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