AI Paper Atlas

Research map / Time series / Foundation models for time series

Revealing Scaling Paradox in Large-scale Time Series Models: Implications for More Efficient and Accurate Forecasting

Xin Qiu, Junlong Tong, Yirong Sun, Yunpu Ma, Xiaoyu Shen

ICML 2026 · Time series › Foundation models for time series · approach: LLM-based 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
scale model from Small to LargeLLM4TS (single-dataset) · avg in-domainMSE: 0.2226 -> 0.2846Hurt
scale model from Small to LargeTSFMs (single-dataset) · avg in-domainMSE: 0.2286 -> 0.2540Hurt
add cross-dataset training with scaleLLM4TS (Small-C to Large-C) · avg in-domainMSE: 0.2367 -> 0.2233Improved
scale TSFMs under cross-dataset learning from Small-C to Large-CTSFMs · avg in-domainMSE: 0.2257 -> 0.2543Hurt
increase training data ratio from 20% to 100%LLM4TS and TSFMs · single-dataset MSEMSE: MSE decreases but gap not narrowedNo clear effect
add critical-layer pruning and fine-tuningLLM4TS-Large · single-dataset in-domainMSE: 0.285 -> 0.219Improved
add critical-layer pruning and fine-tuningLLM4TS-Base · single-dataset in-domainMSE: 0.278 -> 0.226Improved
add critical-layer pruning and fine-tuningLLM4TS-Large-C · cross-dataset in-domainMSE: 0.223 -> 0.226Hurt
add critical-layer pruning and fine-tuningTSFMs-Large-C · cross-dataset out-of-domainMSE: 0.872 -> 0.821Improved
add critical-layer pruning and fine-tuningAll six single-dataset models · single-datasetinference efficiency: 2.7Improved
use critical layers instead of full backbone (ablation)FSCA, CALF, OFA · 5 datasets avgMSE: 0.348 -> 0.329Improved
use critical layers instead of random layers (ablation)FSCA, CALF, OFA · 5 datasets avgMSE: 0.343 -> 0.329Improved

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