Mingtian Tan, Mike A Merrill, Zachary Gottesman, Tim Althoff, David Evans, Thomas Hartvigsen. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Large language models (LLMs) are being applied to time series forecasting. But are language models actually useful for time series? In a series of ablation studies on three recent and popular LLM-based time series forecasting methods, we find that removing the LLM component or replacing it with a basic attention layer does not degrade forecasting performance---in most cases, the results even improve! We also find that despite their significant computational cost, pretrained LLMs do no better than models trained from scratch, do not represent the sequential dependencies in time series, and do not assist in few-shot settings. Additionally, we explore time series encoders and find that patching and attention structures perform similarly to LLM-based forecasters. All resources needed to reproduce our work are available: https://github.com/BennyTMT/LLMsForTimeSeries.

Recent research has demonstrated the capability of behavior signals captured by smartphones and wearables for longitudinal behavior modeling. However, there is a lack of a comprehensive public dataset that serves as an open testbed for fair comparison among algorithms. Moreover, prior studies mainly evaluate algorithms using data from a single population within a short period, without measuring the cross-dataset generalizability of these algorithms. We present the first multi-year passive sensing datasets, containing over 700 user-years and 497 unique users’ data collected from mobile and wearable sensors, together with a wide range of well-being metrics. Our datasets can support multiple cross-dataset evaluations of behavior modeling algorithms’ generalizability across different users and years. As a starting point, we provide the benchmark results of 18 algorithms on the task of depression detection. Our results indicate that both prior depression detection algorithms and domain generalization techniques show potential but need further research to achieve adequate cross-dataset generalizability. We envision our multi-year datasets can support the ML community in developing generalizable longitudinal behavior modeling algorithms.

Data analyses are based on a series of "decision points" including data filtering, feature operationalization and selection, model specification, and parametric assumptions. "Multiverse Analysis" research has shown that a lack of exploration of these decisions can lead to non-robust conclusions based on highly sensitive decision points. Importantly, even if myopic analyses are technically correct, analysts' focus on one set of decision points precludes them from exploring alternate formulations that may produce very different results. Prior work has also shown that analysts' exploration is often limited based on their training, domain, and personal experience. However, supporting analysts in exploring alternative approaches is challenging and typically requires expert feedback that is costly and hard to scale.