ML is All You Need

This repo provides few-sentence summaries of trending ML papers mostly focused on time series forecasting, foundation models and anomaly detection. We skip over dataset descriptions (spoiler: everyone uses ETT) and result sections (surprise: everyone's state-of-the-art). Instead, the focus is on the core technical details auch as architecture, preprocessing, learning objective, and what makes the method unique.

📚 Paper Index

Title	Paper	Code	Publisher	Year	Topics
Learning without training: The implicit dynamics of in-context learning	arXiv		Google	2025	Context, LLM
TiRex: Zero-Shot Forecasting Across Long and Short Horizons with Enhanced In-Context Learning	arXiv	GitHub		2025	FM, Patching, Probabilistic
Time-MoE: Billion-Scale Time Series Foundation Models with Mixture of Experts	arXiv	GitHub		2025	FM, MoE, Multi-Horizon
Aurora: A foundation model for the Earth system	Nature	GitHub	Microsoft	2025	FM, Patching
ChronosX: Adapting Pretrained Time Series Models with Exogenous Variables	arXiv	GitHub	Amazon	2025	Covariates, FM, LLM
Are Language Models Actually Useful for Time Series Forecasting?	arXiv	GitHub		2024	LLM
xLSTM: Extended Long Short-Term Memory	arXiv	GitHub	NXAI	2024	Memory
MOMENT: A Family of Open Time-series Foundation Models	arXiv	GitHub		2024	Anomaly, FM, Patching, Representation
Chronos: Learning the Language of Time Series	arXiv	GitHub	Amazon	2024	FM, LLM, Probabilistic
Titans: Learning to Memorize at Test Time	arXiv		Google	2024	FM, Memory, Static
Lag-Llama: Towards Foundation Models for Probabilistic Time Series Forecasting	arXiv	GitHub		2024	Covariates, FM, Multi-Horizon, Probabilistic
TimesFM: A decoder-only foundation model for time-series forecasting	arXiv		Google	2024	FM, Multi-Horizon, Patching
PatchTST: A Time Series is Worth 64 Words - Long-term Forecasting with Transformers	arXiv	GitHub	IBM	2023	Patching, Representation
N-HiTS: Neural Hierarchical Interpolation for Time Series Forecasting	arXiv	GitHub	Nixtla	2022	Interpretable, Multi-Horizon
Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting	arXiv	GitHub		2021	FM, Multi-Horizon
GDN: Graph Neural Network-Based Anomaly Detection in Multivariate Time Series	arXiv	GitHub		2021	Anomaly
TFT: Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting	arXiv		Google	2020	Covariates, Probabilistic, Static
N-BEATS: Neural Basis Expansion Analysis for Interpretable Time Series Forecasting	arXiv	GitHub	Element AI	2020	Interpretable
USAD: UnSupervised Anomaly Detection on Multivariate Time Series	ACM Digital Library			2020	Anomaly
MTAD-GAT: Multivariate Time-series Anomaly Detection via Graph Attention Network	arXiv	GitHub		2020	Anomaly