{
  "abstract": "Can the spatial organization of cortical activity help transfer frozen image representations to fMRI prediction? We introduce FlatClip, which renders cortical activity as anatomy-linked flatmaps and extracts features with a frozen SigLIP2 encoder. Only a lightweight downstream probe is trained, without fMRI-specific encoder pretraining or fine-tuning. Across four resting-state benchmarks, FlatClip outperforms the evaluated ROI-level baselines on HCP and ADNI, but is weaker on PPMI and below the strongest voxel-level models overall. On NSD visual-fMRI decoding, restricting inputs to visual or NSD-provided task-active cortex improves recognition over whole-cortex inputs. Spatial perturbations reduce performance with both retrained and fixed readouts, and anatomy-linked arrangements outperform cortical-pixel permutations across three colormaps. Together, these results demonstrate a practical route for reusing image-pretrained features in fMRI and highlight cortical arrangement and input-region selection as important design choices. Code is available at https://github.com/OneMore1/FlatClip.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2609.31204",
  "author_total": 7,
  "authors": [
    {
      "name": "Mo Wang"
    },
    {
      "name": "Wenhao Ye"
    },
    {
      "name": "Zihan Ning"
    },
    {
      "name": "Jiayu Zuo"
    },
    {
      "name": "Junfeng Xia"
    },
    {
      "name": "Hongkai Wen"
    },
    {
      "name": "Quanying Liu"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/exzn93j0nm0pg.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2609.31204",
  "fulltext_url": null,
  "id": "exzn93j0nm0pg",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2609.31204",
  "language": null,
  "licence": "CC-BY-NC-SA-4.0",
  "licence_label": "CC BY-NC-SA 4.0",
  "licence_source": "article",
  "licence_url": "http://creativecommons.org/licenses/by-nc-sa/4.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "cs.CE"
    ],
    [
      "arxiv",
      "cs.CV"
    ],
    [
      "arxiv",
      "q-bio.NC"
    ],
    [
      "openalex",
      "2805|Functional Brain Connectivity Studies"
    ]
  ],
  "openalex": "W7214657206",
  "pdf_url": "https://arxiv.org/pdf/2609.31204",
  "pmcid": null,
  "pmid": null,
  "published": "2026-10-15",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/exzn93j0nm0pg.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:29:55Z",
      "native_id": "2609.31204",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2609.31204"
    }
  ],
  "subjects": [
    {
      "field": "biology",
      "origin": "arxiv:q-bio.NC",
      "subfield": "neuroscience"
    }
  ],
  "title": "FlatClip: Reusing Image Foundation Models for fMRI Representation Learning via Cortical Flatmaps",
  "title_text": "FlatClip: Reusing Image Foundation Models for fMRI Representation Learning via Cortical Flatmaps",
  "type": "preprint",
  "updated_at": "2026-10-30T12:29:55Z",
  "url": "https://research.draconitic.com/paper/exzn93j0nm0pg",
  "venue": "arXiv",
  "versions": []
}