Metraipolis › archive › dc_a436f574cb80
STAGE: Subspace-Targeted Affine Generative Erasure for Text-to-3D Models — arXiv:2610.01302
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.01302
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-06T19:13:17.378Z
- SHA-256 of the archived text
e24979891d446be143d21a3debdc3fac625655c4f786169afff794e1d74d7288- Stable URL of this page
https://metraipolis.com/documento/dc_a436f574cb80
How to cite
STAGE: Subspace-Targeted Affine Generative Erasure for Text-to-3D Models — arXiv:2610.01302 — arXiv (en), https://arxiv.org/abs/2610.01302 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a436f574cb80 (Metraipolis archive dc_a436f574cb80, sha256 e24979891d446be143d21a3debdc3fac625655c4f786169afff794e1d74d7288)
BibTeX
@misc{dca436f574cb80,
title = {STAGE: Subspace-Targeted Affine Generative Erasure for Text-to-3D Models — arXiv:2610.01302},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.01302},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a436f574cb80; sha256 e24979891d446be143d21a3debdc3fac625655c4f786169afff794e1d74d7288},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.01302. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Concept erasure suppresses a target concept while preserving behavior on unrelated inputs. Existing closed-form methods were designed for 2D image diffusion and assume a single generative pathway, so one edit must cover geometry and texture at once. Native 3D generators, which synthesize structured 3D representations directly rather than by lifting 2D samples, violate this assumption. We show that shape and object concepts must be erased in the structural stage of the pipeline and material concepts in the appearance stage. We therefore formulate erasure in native text-to-3D as a stage-aware editing problem and introduce STAGE, a training-free, closed-form framework. STAGE confines each edit to the low-dimensional subspace spanned by the differences between erase and anchor embeddings, and relaxes the norm-preserving (orthogonal) constraint of prior editors into a least-squares affine correction that maps target activations onto safe anchors subject to a penalty on the displacement of retained prompts. The correction applies to the structural stage, the appearance stage, or both. We find that the stage an edit must reach is determined by concept type. On TRELLIS, the standard open native 3D generator, across 15 shape, material, and object concepts, STAGE reaches 66.7 on a composite score that balances forgetting the target concept against preserving everything else, aggregating CLIP-based semantic and physical metrics, versus 53.2 for the strongest adapted baseline. Code: this https URL Project Page this https URL Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-10-02 in cs.CV). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2610.01302 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-06T17:37:36.774Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-10-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).