AI Paper Atlas

Research map / Benchmarks / Weather (avg over horizons)

Weather (avg over horizons) leaderboard

10 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on Weather (avg over horizons). Best MSE: 0.186 (KineFlow, ICML 2026).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

Weather (avg over horizons) · MSE

10 papers, each with its best reported number for its own method; lower is better.

#MSE ↓MethodPaperTopic
10.186KineFlowKineFlow: Kinematic Second-Order Flow Matching for Time-Series Forecasting
ICML 2026 · Table 1
Causal and generative time series
20.201IdealTSFIdealTSF: Can Non-Ideal Data Contribute to Enhancing the Performance of Time Series Forecasting Models?
AAAI 2026 · Table 1
Long-term forecasting
30.213TStructTStruct: Learning Shared Temporal Structures for Long-Term Time-Series Forecasting
NeurIPS 2026 · Table 1
Long-term forecasting
40.215FreDNFreDN: Spectral Disentanglement for Time Series Forecasting via Learnable Frequency Decomposition
AAAI 2026 · Table 1
Long-term forecasting
50.217PMDformer
lookback 720
PMDformer: Patch-Mean Decoupling Transformer for Long-term Forecasting
ICLR 2026 · Table 2
Long-term forecasting
60.22CycleNet+ALW
CycleNet
Efficiently Enhancing Long-term Series Forecasting via Adaptive Lookback with Wavelets
AAAI 2026 · Table 2
Long-term forecasting
70.224GraFT
GPT-2 (4 layers)
GraFT: Infusing Pre-trained Transformers with Relational Structure for Time Series Forecasting
AAAI 2026 · Table 1
Long-term forecasting
80.24Dynamic TMoEDynamic TMoE: A Drift-Aware Dynamic Mixture of Experts Framework for Non-Stationary Time Series Forecasting
ICML 2026 · Table 1
Causal and generative time series
90.247InfoGeo
iTransformer
Information Geometry Loss for Time Series Forecasting
ICML 2026 · Table 1
Causal and generative time series
100.254Prime attention
iTransformer
Dynamic Relational Priming Improves Transformer in Multivariate Time Series
ICML 2026 · Table 1
Long-term forecasting

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.