Investigating simple target-covariate relationships for Chronos-2 and TabPFN-TS

Gaspard Berthelier, Mariia Baranova, Andrei-Tiberiu Pantea, Etienne Le Naour, Adrien Petralia, Tahar Nabil, Themis Palpanas

Presentation Attendance: No, we cannot present in-person

Keywords: Time Series Forecasting, Foundation Models, Covariate, Benchmark

TL;DR: We compare Chronos-2 and TabPFN-TS on controlled experiments with simple covariate–target relationships. Despite strong benchmark results for Chronos-2, TabPFN-TS better captures simple covariate–target relationships, especially at short horizons.

Abstract: Time Series Foundation Models (TSFMs) have recently achieved state-of-the-art performance, often outperforming supervised models in zero-shot settings. Recent TSFM architectures, such as Chronos-2 and TabPFN-TS, aim to integrate covariates. In this paper, we design controlled experiments based on simple target–covariate relationships to assess this integration capability. Our results show that TabPFN-TS captures these relationships more effectively than Chronos-2, especially for short horizons, suggesting that the strong benchmark performance of Chronos-2 does not automatically translate into optimal modeling of simple covariate–target dependencies.

Track: Industry and Application Track (max 2 pages)

Submission Number: 107