{
  "id": "2212.07699",
  "version": "v1",
  "published": "2022-12-15T10:20:42.000Z",
  "updated": "2022-12-15T10:20:42.000Z",
  "title": "Retrieval-based Disentanglement with Distant Supervision",
  "authors": [
    "Jiawei Zhou",
    "Xiaoguang Li",
    "Lifeng Shang",
    "Xin Jiang",
    "Qun Liu",
    "Lei Chen"
  ],
  "categories": [
    "cs.CL",
    "cs.AI",
    "cs.CV"
  ],
  "abstract": "Disentangled representation learning remains challenging as ground truth factors of variation do not naturally exist. To address this, we present Vocabulary Disentanglement Retrieval~(VDR), a simple yet effective retrieval-based disentanglement framework that leverages nature language as distant supervision. Our approach is built upon the widely-used bi-encoder architecture with disentanglement heads and is trained on data-text pairs that are readily available on the web or in existing datasets. This makes our approach task- and modality-agnostic with potential for a wide range of downstream applications. We conduct experiments on 16 datasets in both text-to-text and cross-modal scenarios and evaluate VDR in a zero-shot setting. With the incorporation of disentanglement heads and a minor increase in parameters, VDR achieves significant improvements over the base retriever it is built upon, with a 9% higher on NDCG@10 scores in zero-shot text-to-text retrieval and an average of 13% higher recall in cross-modal retrieval. In comparison to other baselines, VDR outperforms them in most tasks, while also improving explainability and efficiency.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2022-12-15T10:20:42.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "distant supervision",
      "representation learning remains challenging",
      "disentanglement heads",
      "vdr achieves significant improvements",
      "ground truth factors"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}