{
  "annotation_id": "92671b99-4975-4ae2-8623-606d4db91066",
  "date_created": "2026-02-17T05:53:19.895000Z",
  "date_modified": "2026-02-17T05:53:19.895000Z",
  "file_hash": "c88c9f4006cd18477695e939e374a67e33885bb7d67c04306078169de4bd6075",
  "private": false,
  "record": {
    "abstract": "Language models perform well on grammatical agreement, but it is unclear whether this reflects rule-based generalization or memorization. We study this question for German definite singular articles, whose forms depend on gender and case. Using GRADIEND, a gradient-based interpretability method, we learn parameter update directions for gender-case specific article transitions. We find that updates learned for a specific gender-case article transition frequently affect unrelated gender-case settings, with substantial overlap among the most affected neurons across settings. These results argue against a strictly rule-based encoding of German definite articles, indicating that models at least partly rely on memorized associations rather than abstract grammatical rules.",
    "arxiv_id": "2601.09313",
    "authors": [
      "Jonathan Drechsel",
      "Erisa Bytyqi",
      "Steffen Herbold"
    ],
    "categories": [
      "cs.CL",
      "cs.AI"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "title": "Understanding or Memorizing? A Case Study of German Definite Articles in Language Models",
    "url": "https://arxiv.org/abs/2601.09313",
    "version": "v1"
  },
  "schema_id": "dorsal/arxiv",
  "source": {
    "execution_id": "9bcfb36c-0383-4b6f-b377-88daba328ef2",
    "id": "arXiv Dataset",
    "type": "Model",
    "variant": "snapshot-2026-01-17",
    "version": "0.1.0"
  },
  "user_id": 1000002
}