AI Paper Atlas

Research map / Time series / Foundation models for time series

CoGenCast: A Coupled Autoregressive–Flow Generative Framework for Time Series Forecasting

Mingyue Cheng, Yaguo Liu, Daoyu Wang, Xiaoyu Tao, Qi Liu

ICML 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 LLM encoder (bidirectional) to the decoder-only variant (ablation)CoGenCast · Energy, Environ., Exchange, WindMSE: Energy 0.953 -> 0.293; Wind 2.010 -> 0.532Improved
add decoder with flow matching on top of encoder-only variant (ablation)CoGenCast · EnergyMSE: 0.331 -> 0.293Improved
add autoregressive mechanism (ablation)CoGenCast · WindMSE: 0.673 -> 0.532Improved
add flow-matching mechanism (ablation)CoGenCast · WindMSE: 0.611 -> 0.532Improved
couple both AR and flow versus neither (ablation)CoGenCast · ETTh1MSE: 0.372 -> 0.328Improved
use average-velocity one-step generator instead of vanilla flow matching (ablation)CoGenCast · ExchangeMSE: 0.051 -> 0.031Improved
use average-velocity generator instead of vanilla diffusion (ablation)CoGenCast · EnergyMSE: 0.325 -> 0.293Improved
increase function evaluations from 1 to 3 (ablation)CoGenCast · Energy, Environ., ETTh1, SolarMSENo clear effect
use linear noise scheduler instead of cosine (ablation)CoGenCast · Energy, Environ., ETTh1MSEImproved
use pretrained Qwen3-0.6B instead of vanilla Transformer backbone (ablation)CoGenCast · WindMSE: 0.587 -> 0.532Improved
scale backbone from Qwen3-0.6B to Qwen3-4B (ablation)CoGenCast · Energy, Environ., Exchange, WindMSE: Wind 0.532 -> 0.525; Energy 0.293 -> 0.290No clear effect
swap Qwen3-0.6B for Llama3.2-1B backbone (ablation)CoGenCast · ETTh2, ETTm1, ETTm2, ExchangeMSE: ETTm1 0.224 -> 0.221No clear effect

Earlier methods this paper compares with

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