Towards Interpretable Time Series Foundation Models

• URL: http://arxiv.org/abs/2507.07439v1
• Date: Thu, 10 Jul 2025 05:29:34 GMT
• Title: Towards Interpretable Time Series Foundation Models
• Authors: Matthieu Boileau, Philippe Helluy, Jeremy Pawlus, Svitlana Vyetrenko,
• Abstract summary: We generate natural language annotations using a large multimodal model and use these to supervise the fine-tuning of compact Qwen models.<n>Our results highlight the feasibility of compressing time series understanding into lightweight, language-capable models suitable for on-device or privacy-sensitive deployment.
• Score: 0.0
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: In this paper, we investigate the distillation of time series reasoning capabilities into small, instruction-tuned language models as a step toward building interpretable time series foundation models. Leveraging a synthetic dataset of mean-reverting time series with systematically varied trends and noise levels, we generate natural language annotations using a large multimodal model and use these to supervise the fine-tuning of compact Qwen models. We introduce evaluation metrics that assess the quality of the distilled reasoning - focusing on trend direction, noise intensity, and extremum localization - and show that the post-trained models acquire meaningful interpretive capabilities. Our results highlight the feasibility of compressing time series understanding into lightweight, language-capable models suitable for on-device or privacy-sensitive deployment. This work contributes a concrete foundation toward developing small, interpretable models that explain temporal patterns in natural language.

Related papers

• The language of time: a language model perspective on time-series foundation models [7.113398204739559]
  We study the representation learning mechanisms and generalization capabilities of patch-based time series foundation models.<n>Our work provides a rigorous theoretical cornerstone for understanding, evaluating, and improving the safety and reliability of large-scale time series foundation models.
  arXiv  Detail & Related papers  (2025-06-29T14:03:34Z)
• Context parroting: A simple but tough-to-beat baseline for foundation models in scientific machine learning [6.445605125467573]
  We show that foundation models applied to physical systems can give accurate predictions, but that they fail to develop meaningful representations of the underlying physics.<n>Instead, foundation models often forecast by context parroting, a simple zero-shot forecasting strategy.<n>We draw a parallel between context parroting and induction heads, which explains why large language models trained on text can be repurposed for time series forecasting.
  arXiv  Detail & Related papers  (2025-05-16T15:14:47Z)
• TimesBERT: A BERT-Style Foundation Model for Time Series Understanding [72.64824086839631]
  GPT-style models have been positioned as foundation models for time series forecasting.<n>BERT-style architecture has not been fully unlocked for time series understanding.<n>We design TimesBERT to learn generic representations of time series.<n>Our model is pre-trained on 260 billion time points across diverse domains.
  arXiv  Detail & Related papers  (2025-02-28T17:14:44Z)
• Enhancing Foundation Models for Time Series Forecasting via Wavelet-based Tokenization [74.3339999119713]
  We develop a wavelet-based tokenizer that allows models to learn complex representations directly in the space of time-localized frequencies.<n>Our method first scales and decomposes the input time series, then thresholds and quantizes the wavelet coefficients, and finally pre-trains an autoregressive model to forecast coefficients for the forecast horizon.
  arXiv  Detail & Related papers  (2024-12-06T18:22:59Z)
• Context is Key: A Benchmark for Forecasting with Essential Textual Information [87.3175915185287]
  "Context is Key" (CiK) is a forecasting benchmark that pairs numerical data with diverse types of carefully crafted textual context.<n>We evaluate a range of approaches, including statistical models, time series foundation models, and LLM-based forecasters.<n>We propose a simple yet effective LLM prompting method that outperforms all other tested methods on our benchmark.
  arXiv  Detail & Related papers  (2024-10-24T17:56:08Z)
• Exploring Representations and Interventions in Time Series Foundation Models [17.224575072056627]
  Time series foundation models (TSFMs) promise to be powerful tools for a wide range of applications.<n>Their internal representations and learned concepts are still not well understood.<n>This study investigates the structure and redundancy of representations across various TSFMs.
  arXiv  Detail & Related papers  (2024-09-19T17:11:27Z)
• A decoder-only foundation model for time-series forecasting [23.824504640087753]
  Our model is based on pretraining a patched-decoder style attention model on a large time-series corpus. It can work well across different forecasting history lengths, prediction lengths and temporal granularities.
  arXiv  Detail & Related papers  (2023-10-14T17:01:37Z)
• Expedited Training of Visual Conditioned Language Generation via Redundancy Reduction [61.16125290912494]
  $textEVL_textGen$ is a framework designed for the pre-training of visually conditioned language generation models. We show that our approach accelerates the training of vision-language models by a factor of 5 without a noticeable impact on overall performance.
  arXiv  Detail & Related papers  (2023-10-05T03:40:06Z)
• Time-LLM: Time Series Forecasting by Reprogramming Large Language Models [110.20279343734548]
  Time series forecasting holds significant importance in many real-world dynamic systems. We present Time-LLM, a reprogramming framework to repurpose large language models for time series forecasting. Time-LLM is a powerful time series learner that outperforms state-of-the-art, specialized forecasting models.
  arXiv  Detail & Related papers  (2023-10-03T01:31:25Z)
• SINC: Self-Supervised In-Context Learning for Vision-Language Tasks [64.44336003123102]
  We propose a framework to enable in-context learning in large language models. A meta-model can learn on self-supervised prompts consisting of tailored demonstrations. Experiments show that SINC outperforms gradient-based methods in various vision-language tasks.
  arXiv  Detail & Related papers  (2023-07-15T08:33:08Z)
• Model Criticism for Long-Form Text Generation [113.13900836015122]
  We apply a statistical tool, model criticism in latent space, to evaluate the high-level structure of generated text. We perform experiments on three representative aspects of high-level discourse -- coherence, coreference, and topicality. We find that transformer-based language models are able to capture topical structures but have a harder time maintaining structural coherence or modeling coreference.
  arXiv  Detail & Related papers  (2022-10-16T04:35:58Z)
• Pitfalls of Static Language Modelling [41.76918612574081]
  We show that state-of-the-art Transformer models perform worse in the realistic setup of predicting future utterances from beyond their training period. We argue that now is the right time to rethink our static language modelling evaluation protocol.
  arXiv  Detail & Related papers  (2021-02-03T09:01:49Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.