{
  "abstract": "Instance-wise feature selection is a valuable tool for interpreting labeled data and the predictions of black-box models. In contrast to global feature selection techniques, instance-wise methods dynamically identify important features for each instance. A growing number of methods learn a selector, which identifies important features, and a predictor, which uses these to make predictions. However, these pioneering methods face challenges including information leakage and lack of differentiability, which can slow training. In this paper, we present Hide&Seek, an end-to-end differentiable model for instance-wise feature selection. We jointly learn feature selection and prediction under a single objective without information leakage. Hide&Seek outperforms existing state-of-the-art models across a range of experiments and is fast to train. We achieve this by reformulating feature removal as a differentiable operation where instead of discretely removing features, we replace a proportion of each feature. Training is further stabilized via a parsimony-weight annealing framework.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": "Journal reference: Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:27939-27965, 2026",
  "arxiv": "2608.16689",
  "author_total": 3,
  "authors": [
    {
      "name": "Tal Ellinson"
    },
    {
      "name": "Hadi Mohasel Afshar"
    },
    {
      "name": "Sally Cripps"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/ezj7dgs420628.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2608.16689",
  "fulltext_url": null,
  "id": "ezj7dgs420628",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2608.16689",
  "language": null,
  "licence": "CC-BY-4.0",
  "licence_label": "CC BY 4.0",
  "licence_source": "article",
  "licence_url": "http://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "cs.LG"
    ],
    [
      "arxiv",
      "stat.ML"
    ],
    [
      "openalex",
      "1702|Explainable Artificial Intelligence (XAI)"
    ]
  ],
  "openalex": "W7203711012",
  "pdf_url": "https://arxiv.org/pdf/2608.16689",
  "pmcid": null,
  "pmid": null,
  "published": "2026-08-17",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/ezj7dgs420628.ris",
  "sources": [
    {
      "first_seen": "2026-11-02T12:21:25Z",
      "native_id": "2608.16689",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2608.16689"
    }
  ],
  "subjects": [
    {
      "field": "mathematics",
      "origin": "arxiv:stat.ML",
      "subfield": "statistics"
    }
  ],
  "title": "Hide&Seek: Learning to Explain in an End-to-End Differentiable Network",
  "title_text": "Hide&Seek: Learning to Explain in an End-to-End Differentiable Network",
  "type": "preprint",
  "updated_at": "2026-11-04T15:20:53Z",
  "url": "https://research.draconitic.com/paper/ezj7dgs420628",
  "venue": "arXiv",
  "versions": []
}