{
  "abstract": "Real-world time series exhibit complex dynamics across multiple temporal resolutions, yet effectively incorporating multi-scale information into generative modeling remains challenging without increasing generative complexity. We propose UniFlow, a unified framework that integrates multi-scale representation learning with efficient non-autoregressive generation through a shared-semantic discrete interface. Specifically, a multi-scale tokenizer employs shared-index voting to consolidate complementary temporal information into a single public-ID sequence, while scale-specific codebooks preserve the information required for reconstruction. Building on this unified representation, we develop geometry-aware categorical flow matching, which initializes transport from learned low-rank anchors and generates public identities non-autoregressively. This design captures multi-scale temporal structure without requiring multiple generative token streams or sequential token decoding. Experiments across four benchmarks and both short- and long-term settings demonstrate improved generative fidelity, strong cross-scale consistency, and substantial inference speedups over representative autoregressive and diffusion-based baselines. Code is available at https://github.com/William-Liwei/uniflow",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2611.04825",
  "author_total": 5,
  "authors": [
    {
      "name": "Shibo Feng"
    },
    {
      "name": "Wei Li"
    },
    {
      "name": "Min Wu"
    },
    {
      "name": "Chunyan Miao"
    },
    {
      "name": "Peilin Zhao"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/9mg0e8ngyh29q.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2611.04825",
  "fulltext_url": null,
  "id": "9mg0e8ngyh29q",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2611.04825",
  "language": null,
  "licence": "LicenseRef-arXiv-nonexclusive-1.0",
  "licence_label": "arXiv licence",
  "licence_source": "article",
  "licence_url": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "cs.AI"
    ],
    [
      "arxiv",
      "stat.ML"
    ]
  ],
  "openalex": null,
  "pdf_url": "https://arxiv.org/pdf/2611.04825",
  "pmcid": null,
  "pmid": null,
  "published": "2026-11-03",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/9mg0e8ngyh29q.ris",
  "sources": [
    {
      "first_seen": "2026-11-04T15:21:34Z",
      "native_id": "2611.04825",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2611.04825"
    }
  ],
  "subjects": [
    {
      "field": "mathematics",
      "origin": "arxiv:stat.ML",
      "subfield": "statistics"
    }
  ],
  "title": "UniFlow: Unified Multi-Scale Flow Matching for Time Series Generation",
  "title_text": "UniFlow: Unified Multi-Scale Flow Matching for Time Series Generation",
  "type": "preprint",
  "updated_at": "2026-11-06T18:21:10Z",
  "url": "https://research.draconitic.com/paper/9mg0e8ngyh29q",
  "venue": "arXiv",
  "versions": []
}