{"ID":2891308,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2508.00881","arxiv_id":"2508.00881","title":"Hallucination Detection and Mitigation with Diffusion in Multi-Variate Time-Series Foundation Models","abstract":"Foundation models for natural language processing have many coherent definitions of hallucination and methods for its detection and mitigation. However, analogous definitions and methods do not exist for multi-variate time-series (MVTS) foundation models. We propose new definitions for MVTS hallucination, along with new detection and mitigation methods using a diffusion model to estimate hallucination levels. We derive relational datasets from popular time-series datasets to benchmark these relational hallucination levels. Using these definitions and models, we find that open-source pre-trained MVTS imputation foundation models relationally hallucinate on average up to 59.5% as much as a weak baseline. The proposed mitigation method reduces this by up to 47.7% for these models. The definition and methods may improve adoption and safe usage of MVTS foundation models.","short_abstract":"Foundation models for natural language processing have many coherent definitions of hallucination and methods for its detection and mitigation. However, analogous definitions and methods do not exist for multi-variate time-series (MVTS) foundation models. We propose new definitions for MVTS hallucination, along with ne...","url_abs":"https://arxiv.org/abs/2508.00881","url_pdf":"https://arxiv.org/pdf/2508.00881v1","authors":"[\"Vijja Wichitwechkarn\",\"Charles Fox\",\"Ruchi Choudhary\"]","published":"2025-07-23T11:09:19Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cs.CL\"]","methods":"[\"Diffusion Model\"]","has_code":false}