{
  "absUrl": "https://arxiv.org/abs/2609.27901",
  "arxivId": "2609.27901",
  "arxivVersion": null,
  "arxivVersionedId": null,
  "authors": [
    {
      "affiliations": [
        "Bar-Ilan University"
      ],
      "name": "Ohad Rahamim"
    },
    {
      "affiliations": [
        "NVIDIA"
      ],
      "name": "Dvir Samuel"
    },
    {
      "affiliations": [
        "Bar-Ilan University"
      ],
      "name": "Idan Schwartz"
    },
    {
      "affiliations": [
        "Bar-Ilan University",
        "NVIDIA"
      ],
      "name": "Gal Chechik"
    }
  ],
  "contract": "researcher-sidecars-v1",
  "id": "2609.27901",
  "originalTitle": "All modalities are equal, but video is more equal: Closing the Cross-Attention Gap in Joint Video Generation",
  "pdfUrl": "https://arxiv.org/pdf/2609.27901.pdf",
  "schemaVersion": 1,
  "type": "preprint"
}