AI TrackerFrontier desk · AEST
Back to news
normalaitech

Quantifying the Sampling Frequency Invariance of Time-Series Foundation Models for Sensor Data Applications

arXiv cs.AI · 11 Nov 2026, 04:01 AEST

Time-Series Foundation Models (TSFMs) are increasingly utilized in sensor applications with varying sampling frequencies. This study stress-tests four prominent TSFMs to quantify their sampling frequency invariance and its impact on performance.

Key facts

Read aloud

Time-Series Foundation Models TSFMs are increasingly utilized in sensor applications with varying sampling frequencies. This study stress-tests four prominent TSFMs to quantify their sampling frequency invariance and its impact on performance. Time-Series Foundation Models TSFMs are increasingly used in sensor applications, where sampling frequency SF varies across sensors, devices, and operating modes e.g., power saving. In this paper, we stress-test four prominent TSFMs: MOMENT, Chronos-two, Moirai-MoE, and TimesFM, to quantify their SF-invariance and investigate whether their behavior tracks the resulting loss of information bandwidth rather than the SF value itself. Our results characterize: i how SF-invariance depends on the margin between signal's native-SF and the bandwidth its task requires, ii how much of that invariance the choice of TSFM explains, and iii how much of TSFM's performance degradation under a lower SF reflects information loss rather than a limitation of their abilities.

Sources in this cluster