Cover Image

Why Do Transformers Fail to Forecast Time Series In-Context?

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Zhou, Yufa, Wang, Yixiao, Goel, Surbhi, Zhang, Anru R.
Format:	Preprint
Published:	2025
Subjects:	Machine Learning Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

The Geometry of Reasoning: Flowing Logics in Representation Space
by: Zhou, Yufa, et al.
Published: (2025)

Curse of Attention: A Kernel-Based Perspective for Why Transformers Fail to Generalize on Time Series Forecasting and Beyond
by: Ke, Yekun, et al.
Published: (2024)

Why Do Time Series Models Need Long Context Windows?
by: Butera, Luca, et al.
Published: (2026)

When Will It Fail?: Anomaly to Prompt for Forecasting Future Anomalies in Time Series
by: Park, Min-Yeong, et al.
Published: (2025)

Decentralized Attention Fails Centralized Signals: Rethinking Transformers for Medical Time Series
by: Yu, Guoqi, et al.
Published: (2026)

Breaking the Context Bottleneck on Long Time Series Forecasting
by: Ma, Chao, et al.
Published: (2024)

HTMformer: Hybrid Time and Multivariate Transformer for Time Series Forecasting
by: Wang, Tan, et al.
Published: (2025)

TimeXer: Empowering Transformers for Time Series Forecasting with Exogenous Variables
by: Wang, Yuxuan, et al.
Published: (2024)

Context-Aware Probabilistic Modeling with LLM for Multimodal Time Series Forecasting
by: Yao, Yueyang, et al.
Published: (2025)

VARMA-Enhanced Transformer for Time Series Forecasting
by: Song, Jiajun, et al.
Published: (2025)

Multi-Layer Transformers Gradient Can be Approximated in Almost Linear Time
by: Liang, Yingyu, et al.
Published: (2024)

SEDformer: Event-Synchronous Spiking Transformers for Irregular Telemetry Time Series Forecasting
by: Zhou, Ziyu, et al.
Published: (2026)

Why Low-Precision Transformer Training Fails: An Analysis on Flash Attention
by: Qiu, Haiquan, et al.
Published: (2025)

PSformer: Parameter-efficient Transformer with Segment Attention for Time Series Forecasting
by: Wang, Yanlong, et al.
Published: (2024)

L-Drive: Beyond a Single Mapping-Latent Context Drives Time Series Forecasting
by: Zhang, Fan, et al.
Published: (2026)

Context Matters: Leveraging Contextual Features for Time Series Forecasting
by: Chattopadhyay, Sameep, et al.
Published: (2024)

Abnormality Forecasting: Time Series Anomaly Prediction via Future Context Modeling
by: Zhao, Sinong, et al.
Published: (2024)

CITRAS: Covariate-Informed Transformer for Time Series Forecasting
by: Yamaguchi, Yosuke, et al.
Published: (2025)

Fredformer: Frequency Debiased Transformer for Time Series Forecasting
by: Piao, Xihao, et al.
Published: (2024)

LATST: Are Transformers Necessarily Complex for Time-Series Forecasting
by: Liang, Dizhen
Published: (2024)

Unified Training of Universal Time Series Forecasting Transformers
by: Woo, Gerald, et al.
Published: (2024)

Enhancing Large Language Models for Time-Series Forecasting via Vector-Injected In-Context Learning
by: Zhang, Jianqi, et al.
Published: (2026)

Ellipsoidal Time Series Forecasting
by: Wang, Qilin
Published: (2025)

DSAT-HD: Dual-Stream Adaptive Transformer with Hybrid Decomposition for Multivariate Time Series Forecasting
by: Wang, Zixu, et al.
Published: (2025)

TimeMixer: Decomposable Multiscale Mixing for Time Series Forecasting
by: Wang, Shiyu, et al.
Published: (2024)

Attention as Robust Representation for Time Series Forecasting
by: Niu, PeiSong, et al.
Published: (2024)

Time Series Forecastability Measures
by: Wang, Rui, et al.
Published: (2025)

Considering Nonstationary within Multivariate Time Series with Variational Hierarchical Transformer for Forecasting
by: Wang, Muyao, et al.
Published: (2024)

HDT: Hierarchical Discrete Transformer for Multivariate Time Series Forecasting
by: Feng, Shibo, et al.
Published: (2025)

Nested Spatio-Temporal Time Series Forecasting
by: Ai, Yinghao, et al.
Published: (2026)

Ister: Linear Transformer for Efficient Multivariate Time Series Forecasting
by: Cao, Fanpu, et al.
Published: (2024)

From Values to Tokens: An LLM-Driven Framework for Context-aware Time Series Forecasting via Symbolic Discretization
by: Tao, Xiaoyu, et al.
Published: (2025)

Wavelet Mixture of Experts for Time Series Forecasting
by: Zhou, Zheng, et al.
Published: (2025)

Enhancing Transformer-based models for Long Sequence Time Series Forecasting via Structured Matrix
by: Zhang, Zhicheng, et al.
Published: (2024)

TripCast: Pre-training of Masked 2D Transformers for Trip Time Series Forecasting
by: Liao, Yuhua, et al.
Published: (2024)

Time-TK: A Multi-Offset Temporal Interaction Framework Combining Transformer and Kolmogorov-Arnold Networks for Time Series Forecasting
by: Zhang, Fan, et al.
Published: (2026)

Calibration of Time-Series Forecasting: Detecting and Adapting Context-Driven Distribution Shift
by: Chen, Mouxiang, et al.
Published: (2023)

Context Neural Networks: A Scalable Multivariate Model for Time Series Forecasting
by: Sriramulu, Abishek, et al.
Published: (2024)

The Forecast After the Forecast: A Post-Processing Shift in Time Series
by: Liang, Daojun, et al.
Published: (2026)

Baguan-TS: A Sequence-Native In-Context Learning Model for Time Series Forecasting with Covariates
by: Yang, Linxiao, et al.
Published: (2026)