2601.10274v1.pdf

{
  "dorsal/arxiv": [
    {
      "date_modified": "2026-02-17T05:53:23.338000Z",
      "hash": "a5aeb497ec4abcf97c47990978e4d4fe9e031952759d0d63dba1a7998f77bfc9",
      "id": "33a28302-eb65-4264-abcc-158ca48c95a9",
      "source": {
        "execution_id": "1fb644e6-8d72-45c8-ab5e-05899e3e4ac6",
        "id": "arXiv Dataset",
        "type": "Model",
        "variant": "snapshot-2026-01-17",
        "version": "0.1.0"
      },
      "user_id": 1000002
    }
  ],
  "file/base": {
    "record": {
      "extension": ".pdf",
      "hash": "a5aeb497ec4abcf97c47990978e4d4fe9e031952759d0d63dba1a7998f77bfc9",
      "media_type": "application/pdf",
      "media_type_prefix": "application",
      "name": "2601.10274v1.pdf",
      "size": 661366
    },
    "source": {
      "execution_id": "c3db5754-076b-4837-9aad-fd537faa8bd5",
      "id": "file/base",
      "type": "Model",
      "version": "1.0.0"
    }
  },
  "file/pdf": {
    "private": false,
    "record": {
      "keywords": [],
      "page_count": 8,
      "producer": "pikepdf 8.15.1",
      "title": "Queueing-Aware Optimization of Reasoning Tokens for Accuracy-Latency Trade-offs in LLM Servers",
      "version": "1.7"
    },
    "source": {
      "execution_id": "6bbc5e88-a140-4202-aa27-63d386ac0d67",
      "id": "dorsal/pdf",
      "type": "Model",
      "version": "1.1.0"
    }
  }
}