{
  "abstract": "Accurate watch-time (WT) prediction is an important requirement for short-video recommendations. Yet WT distributions are near-zero-inflated, long-tailed and multimodal. The recent Exponential-Gaussian Mixture Network (EGMN) models the full conditional WT distribution rather than a single point estimate and achieves state-of-the-art performance. Our large-scale reproduction study reveals that EGMN is vulnerable to variance collapse, component redundancy, and inactive components. We propose a Hierarchical Exponential-Gaussian Mixture (HEGM) model that addresses these failure modes through a hierarchical skip-watch decomposition, structured initialization, removal of the forced Gaussian shift and entropy regularizer and optional KL-based variance regularization for over-parameterized mixtures. Across large-scale industrial and public datasets, under a fixed-backbone head-replacement protocol, HEGM improves ranking accuracy and threshold-event prediction over reproducible lightweight baselines, while maintaining competitive point-estimation accuracy and substantially improving mixture stability and interpretability. A 1.5-month production A/B test confirms statistically significant engagement lifts. Our code and models are publicly released at https://github.com/rw404/HEGM.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2608.23356",
  "author_total": 4,
  "authors": [
    {
      "name": "Sofia Gulevskaia"
    },
    {
      "name": "Mikhail Trapeznikov"
    },
    {
      "name": "Aleksandr Poslavsky"
    },
    {
      "name": "Alexander D'yakonov"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/497hzjrpkgz6n.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2608.23356",
  "fulltext_url": null,
  "id": "497hzjrpkgz6n",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2608.23356",
  "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.IR"
    ],
    [
      "arxiv",
      "cs.LG"
    ],
    [
      "arxiv",
      "stat.ML"
    ],
    [
      "openalex",
      "1710|Recommender Systems and Techniques"
    ]
  ],
  "openalex": "W7204156833",
  "pdf_url": "https://arxiv.org/pdf/2608.23356",
  "pmcid": null,
  "pmid": null,
  "published": "2026-10-21",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/497hzjrpkgz6n.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:34:29Z",
      "native_id": "2608.23356",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2608.23356"
    }
  ],
  "subjects": [
    {
      "field": "mathematics",
      "origin": "arxiv:stat.ML",
      "subfield": "statistics"
    }
  ],
  "title": "Hierarchical Exponential-Gaussian Mixtures for Watch-Time Distribution Prediction",
  "title_text": "Hierarchical Exponential-Gaussian Mixtures for Watch-Time Distribution Prediction",
  "type": "preprint",
  "updated_at": "2026-10-30T12:34:29Z",
  "url": "https://research.draconitic.com/paper/497hzjrpkgz6n",
  "venue": "arXiv",
  "versions": []
}