AI Paper Atlas

Research map / Time series / Long-term forecasting

Understanding Transformers for Time Series: Rank Structure, Flow-of-ranks, and Compressibility

Annan Yu, Danielle C. Maddix, Boran Han, Xiyuan Zhang, Abdul Fatir Ansari, Oleksandr Shchur, Christos Faloutsos, Andrew Gordon Wilson, Michael W. Mahoney, Bernie Wang

ICLR 2026 · Time series › Long-term forecasting · approach: Spatiotemporal forecasting models · 11 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
apply truncated SVD to Chronos attention matrices down to about 23.7% of original size without finetuning (ablation)pretrained Chronos · In-domain / zero-shot forecastingWQL, MASE relative to original: WQL 1.053, MASE 1.005 at ratio 0.237No clear effect
compress Chronos attention below about 20% of original size (ablation)pretrained Chronos · In-domain forecastingWQL: 1.053 -> 1.991 -> 4.409 (ratio 0.237, 0.151, 0.073)Hurt
apply same truncated SVD compression to T5 LLM instead of Chronos (ablation)T5 · language modeling perplexityLPPL relative to original: 1.652 at ratio 0.237; 1.028 at ratio 0.889Hurt
pretrain a compressed model with layer-dependent rank rather than compress a pretrained oneChronos · time-series forecasting Pareto frontieraccuracy vs timeImproved
compress Chronos with low-rank attention and layer-dependent rankChronos · inferenceinference time / memory: -65.4% time, -81.4% memoryImproved
raise the rank of the fixed input embedding from 1 to 32 (ablation)Chronos pretraining · attention matrix numerical ranknumerical rank of W_QImproved
increase patch size k in Chronos-Bolt (ablation)Chronos-Bolt embedding · embedding spectrumnumerical rank of embeddingImproved

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