# ChronosX: Adapting Pretrained Time Series Models with Exogenous Variables

arXiv: https://arxiv.org/abs/2503.12107 · HTML: https://mlsanity.com/paper/2503.12107/

## Key facts

- Cited by 23 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 15 Mar 2025.

## Built on by (most cited first, 23 of 23)

- [fev-bench: A Realistic Benchmark for Time Series Forecasting](https://mlsanity.com/paper/2509.26468/) (2025)
- [Timer-S1: A Billion-Scale Time Series Foundation Model with Serial Scaling](https://mlsanity.com/paper/2603.04791/) (2026)
- [TiRex-2: Generalizing TiRex to Multivariate Data and Streaming](https://mlsanity.com/paper/2607.01204/) (2026)
- [UniCA: Unified Covariate Adaptation for Time Series Foundation Model](https://mlsanity.com/paper/2506.22039/) (2025)
- [CoRA: Covariate-Aware Adaptation of Time Series Foundation Models](https://mlsanity.com/paper/2510.12681/) (2025)
- [CITRAS: Covariate-Informed Transformer for Time Series Forecasting](https://mlsanity.com/paper/2503.24007/) (2025)
- [Post-Training in Time Series Foundation Models: A Unifying Framework](https://mlsanity.com/paper/2607.20002/) (2026)
- [Time-Aware Prior Fitted Networks for Zero-Shot Forecasting with Exogenous Variables](https://mlsanity.com/paper/2603.15802/) (2026)
- [Retrieval Mechanisms Surpass Long-Context Scaling in Time Series Forecasting](https://mlsanity.com/paper/2605.08217/) (2026)
- [Probabilistic Low-Voltage Peak Load Forecasting with Time Series Foundation Models Evaluated on Application-Oriented Metrics](https://mlsanity.com/paper/2607.01966/) (2026)
- [t0: A Time-Series Foundation Model for Forecasting with Context](https://mlsanity.com/paper/2609.24559/) (2026)
- [EventCast: Hybrid Demand Forecasting in E-Commerce with LLM-Based Event Knowledge](https://mlsanity.com/paper/2602.07695/) (2026)
- [Forecast-Then-Optimize Deep Learning Methods](https://mlsanity.com/paper/2506.13036/) (2025)
- [Select, then Balance: Exploring Exogenous Variable Modeling of Spatio-Temporal Forecasting](https://mlsanity.com/paper/2509.05779/) (2025)
- [Evaluating Spatio-Temporal Forecasting Trade-offs Between Graph Neural Networks and Foundation Models](https://mlsanity.com/paper/2511.05179/) (2025)
- [Lightweight Time Series Data Valuation on Time Series Foundation Models via In-Context Finetuning](https://mlsanity.com/paper/2511.11648/) (2025)
- [LightGTS-Cov: Covariate-Enhanced Time Series Forecasting](https://mlsanity.com/paper/2602.10412/) (2026)
- [Assessing the Performance-Efficiency Trade-off of Foundation Models in Probabilistic Electricity Price Forecasting](https://mlsanity.com/paper/2604.14739/) (2026)
- [TFGformer: Multivariate Time Series Forecasting via Time-Frequency Graph Learning and Covariate Fusion](https://mlsanity.com/paper/2607.29459/) (2026)
- [Optimal Stopping of Self-Refining Foundation Models](https://mlsanity.com/paper/2608.10729/) (2026)
- [On the Divergence of Accuracy and Mechanism Consistency in Time Series World Models](https://mlsanity.com/paper/2610.01842/) (2026)
- [BeliefGraph-JEPA: Structured Latent World Models for Action-Conditioned Time Series](https://mlsanity.com/paper/2610.05409/) (2026)
- [ECHO: Correcting Time Series Foundation Model Forecasts with Historical Analogues and Errors](https://mlsanity.com/paper/2610.16433/) (2026)

## Abstract (arXiv)

Covariates provide valuable information on external factors that influence time series and are critical in many real-world time series forecasting tasks. For example, in retail, covariates may indicate promotions or peak dates such as holiday seasons that heavily influence demand forecasts. Recent advances in pretraining large language model architectures for time series forecasting have led to highly accurate forecasters. However, the majority of these models do not readily use covariates as they are often specific to a certain task or domain. This paper introduces a new method to incorporate covariates into pretrained time series forecasting models. Our proposed approach incorporates covariate information into pretrained forecasting models through modular blocks that inject past and future covariate information, without necessarily modifying the pretrained model in consideration. In order to evaluate our approach, we introduce a benchmark composed of 32 different synthetic datasets with varying dynamics to evaluate the effectivity of forecasting models with covariates. Extensive evaluations on both synthetic and real datasets show that our approach effectively incorporates covariate information into pretrained models, outperforming existing baselines.