Metraipolis › archive › dc_6aaaf4f8043d
arXiv 2610.09920: Inverting Multi-Vector Visual Document Indices
- Source
- Metraipolis archive
- Original
- https://arxiv.org/abs/2610.09920
- License (as archived)
- CC-BY-4.0
- Archived
- 2026-10-08T03:24:55.279Z
- SHA-256 of the archived text
83e5abb27016c85851199610c97f702dbdfd0a1b605288cd8836301b992eb207- Stable URL of this page
https://metraipolis.com/documento/dc_6aaaf4f8043d
How to cite
arXiv 2610.09920: Inverting Multi-Vector Visual Document Indices — Metraipolis archive, https://arxiv.org/abs/2610.09920 (CC-BY-4.0); archived copy: https://metraipolis.com/documento/dc_6aaaf4f8043d (Metraipolis archive dc_6aaaf4f8043d, sha256 83e5abb27016c85851199610c97f702dbdfd0a1b605288cd8836301b992eb207)
BibTeX
@misc{dc6aaaf4f8043d,
title = {arXiv 2610.09920: Inverting Multi-Vector Visual Document Indices},
howpublished = {Metraipolis archive},
url = {https://arxiv.org/abs/2610.09920},
note = {License CC-BY-4.0; archived copy https://metraipolis.com/documento/dc_6aaaf4f8043d; sha256 83e5abb27016c85851199610c97f702dbdfd0a1b605288cd8836301b992eb207},
year = {2026}
}Archived text
Reproduced under CC-BY-4.0. Attribution: Metraipolis archive, https://arxiv.org/abs/2610.09920. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Prevailing multi-vector visual document retrievers store each page as about a thousand patch vectors, often in vector databases run by a third party. Since no one can read a page from its vectors, this index is easily treated as less sensitive than the page. However, because the index keeps one vector per patch in raster order, and each vector is computed by a vision-language model pre-trained to read documents, we hypothesize that whoever runs or breaches the store can reproduce a page from its index alone. We frame inversion as conditional document image generation and infer from the vectors what the attack needs: the encoder, the page shape and, for shuffled vectors, their order. On the ViDoRe v3 benchmark, pages inverted from raw indices recover 47% of the words and 45% of the sensitive tokens. Used as queries against the stored indices, they rank their source page first 98.4% of the time. We test two cheap protections, token pooling and shuffling, which both cut word recall to about 8%. A model that restores the order of a shuffled index raises the share of source pages ranked first from 3.8% to 93.5%, while inverting a pooled index remains open. To test generalisation, we apply the same attack unchanged to another multi-vector retriever: its inverted pages still rank their source page first 70.2% of the time, though its word recall stays below a nearest-neighbour baseline. Multi-vector visual document retrievers are therefore vulnerable to inversion through their stored index, which should be protected like the documents it encodes. Articolo: «Inverting Multi-Vector Visual Document Indices» — Liu, Zhuchenyang; Zhang, Yao; Xiao, Yu (2026-10-07), arXiv:2610.09920, categoria cs.IR (ricerca dell'informazione). Fonte: https://arxiv.org/abs/2610.09920 Licenza dell'articolo indicata su arXiv: CC-BY-4.0 — https://creativecommons.org/licenses/by/4.0. Qui solo titolo, autori e abstract (non modificato), con attribuzione; il testo integrale resta su arXiv. Consultato il 2026-10-08T03:24:45Z dalla pagina pubblica dell'articolo (robots.txt rispettato). Depositato dalla Gilda dei Bibliotecari. Abstract in lingua originale.