mlsanityData to 7 Oct 2026

TSFMAudit: Data Contamination Auditing in Forecasting Time Series Foundation Models

arXiv 2605.26161 ↗First posted 24 May 2026cs.LG cs.AIcited by 3

Not on the map. Why?

Abstract (arXiv)

Time series foundation models (TSFMs) are increasingly pretrained on large corpora, raising concerns that evaluation datasets may have been exposed during pretraining and thus yield overly optimistic performance estimates. Auditing such contamination is challenging in time series because signals are continuous and heterogeneous, and often lack corpus documentation. To the best of our knowledge, this is the first work to study pretraining contamination auditing for TSFMs. We formalize the problem of pretraining contamination auditing for TSFMs and propose TSFMAudit, a method based on probe adaptation dynamics. Our key intuition is that contamination manifests as unusually efficient adaptation: after a fine tuning probe, contaminated datasets tend to exhibit faster loss reduction with smaller backbone movement. We evaluate TSFMAudit on 6 TSFMs and 187 datasets using documented training source evidence as supervision, and compare against 10 competitive baselines adapted from the LLM literature.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 39 of its 45 references that are in the map. First posted on arXiv 24 May 2026.

Builds on

39 of 45 references.

References · 39

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2023GPT-4 Technical Reportcited by 18,132
  3. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,221
  4. 2020Extracting Training Data from Large Language Modelscited by 2,144
  5. 2016Understanding deep learning requires rethinking generalizationcited by 1,657
  6. 2021Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecastingcited by 1,654
  7. 2021Membership Inference Attacks From First Principlescited by 839
  8. 2022Quantifying Memorization Across Neural Language Modelscited by 755
  9. 2024Chronos: Learning the Language of Time Seriescited by 712
  10. 2023A decoder-only foundation model for time-series forecastingcited by 650
  11. 2024Unified Training of Universal Time Series Forecasting Transformerscited by 588
  12. 2020What Neural Networks Memorize and Why: Discovering the Long Tail via Influence Estimationcited by 492
27 more
  1. 2024MOMENT: A Family of Open Time-series Foundation Modelscited by 453
  2. 2020Dataset Cartography: Mapping and Diagnosing Datasets with Training Dynamicscited by 423
  3. 2023Detecting Pretraining Data from Large Language Modelscited by 375
  4. 2023NLP Evaluation in trouble: On the Need to Measure LLM Data Contamination for each Benchmarkcited by 339
  5. 2021Monash Time Series Forecasting Archivecited by 256
  6. 2024TFB: Towards Comprehensive and Fair Benchmarking of Time Series Forecasting Methodscited by 242
  7. 2023Proving Test Set Contamination in Black Box Language Modelscited by 227
  8. 2023Rethinking Benchmark and Contamination for Language Models with Rephrased Samplescited by 192
  9. 2024GIFT-Eval: A Benchmark For General Time Series Forecasting Model Evaluationcited by 188
  10. 2024Benchmark Data Contamination of Large Language Models: A Surveycited by 138
  11. 2024Generalization or Memorization: Data Contamination and Trustworthy Evaluation for Large Language Modelscited by 119
  12. 2020The Early Phase of Neural Network Trainingcited by 114
  13. 2025TiRex: Zero-Shot Forecasting Across Long and Short Horizons with Enhanced In-Context Learningcited by 107
  14. 2022Membership Inference Attacks by Exploiting Loss Trajectorycited by 93
  15. 2024Investigating Data Contamination for Pre-training Language Modelscited by 91
  16. 2024VisionTS: Visual Masked Autoencoders Are Free-Lunch Zero-Shot Time Series Forecasterscited by 87
  17. 2025Moirai 2.0: When Less Is More for Time Series Forecastingcited by 59
  18. 2024TSFM-Bench: A Comprehensive and Unified Benchmark of Foundation Models for Time Series Forecastingcited by 55
  19. 2024ConStat: Performance-Based Contamination Detection in Large Language Modelscited by 33
  20. 2026It's TIME: Towards the Next Generation of Time Series Forecasting Benchmarkscited by 25
  21. 2025MIRA: Medical Time Series Foundation Model for Real-World Health Datacited by 21
  22. 2025Exploring the limits of strong membership inference attacks on large language modelscited by 20
  23. 2025How Contaminated Is Your Benchmark? Quantifying Dataset Leakage in Large Language Models with Kernel Divergencecited by 13
  24. 2025CauKer: Classification Time Series Foundation Models Can Be Pretrained on Synthetic Datacited by 12
  25. 2025Rethinking Evaluation in the Era of Time Series Foundation Models: (Un)known Information Leakage Challengescited by 10
  26. 2025Re(Visiting) Time Series Foundation Models in Financecited by 5
  27. 2026Time Series Foundation Models as Strong Baselines in Transportation Forecasting: A Large-Scale Benchmark Analysiscited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026LLM-based Agents for Forecasting and Prediction: Methods, Training, Evaluation, and Applicationscited by 0
  2. 2026A Later Test Set Is Not a New Domain: Pretraining Familiarity Survives a Contamination-Free Hold-Outcited by 0
  3. 2026Evaluation Choices Decide the Forecasting Leaderboard: Evidence from a Production Marketplace Panelcited by 0