Metraipolis › archive › dc_84899a3aa2eb

Low-Rank Prompt Learning for Vision-Language Models with Fixed-Token Bases — arXiv:2609.09462

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.09462
License (as archived)
CC0-1.0
Archived
2026-10-16T10:57:09.680Z
SHA-256 of the archived text
44515682a0644f3716502acfd07f51afcdf7f40f671095b841c886f163fd03d4
Stable URL of this page
https://metraipolis.com/documento/dc_84899a3aa2eb

How to cite

Low-Rank Prompt Learning for Vision-Language Models with Fixed-Token Bases — arXiv:2609.09462 — arXiv (en), https://arxiv.org/abs/2609.09462 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_84899a3aa2eb (Metraipolis archive dc_84899a3aa2eb, sha256 44515682a0644f3716502acfd07f51afcdf7f40f671095b841c886f163fd03d4)
BibTeX
@misc{dc84899a3aa2eb,
  title = {Low-Rank Prompt Learning for Vision-Language Models with Fixed-Token Bases — arXiv:2609.09462},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.09462},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_84899a3aa2eb; sha256 44515682a0644f3716502acfd07f51afcdf7f40f671095b841c886f163fd03d4},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.09462. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Prompt learning adapts CLIP to downstream recognition by replacing hand-written templates with learned continuous context vectors, which in Context Optimization (CoOp) form a dense prompt matrix $\mathbf{P}\in\mathbb{R}^{m\times d}$ trained from only a few examples per class. We study whether this matrix is over-parameterized by factorizing it as $\mathbf{P}=\mathbf{B}\mathbf{A}$, which cuts the trainable prompt parameters from $md$ to $r(m+d)$, and to $rd$ once the token-side factor $\mathbf{B}$ is fixed. Across seven few-shot benchmarks and two CLIP backbones, low-rank prompts match or improve dense CoOp at far fewer parameters, with the clearest gains on low-shot base-to-new generalization. We then find that the token-side factor need not be learned at all: fixing $\mathbf{B}$ to a Gaussian, orthogonal, SVD-derived, or even random basis and training only the embedding-side factor $\mathbf{A}$ stays on par with the fully trainable factorization, and a source-trained $\mathbf{B}$ offers no advantage over a random one. A prompt-factor asymmetry and a local update-space dimension gap show why fixing $\mathbf{B}$ is far less restrictive than fixing $\mathbf{A}$, and a smoothness-only guarantee certifies that optimizing $\mathbf{A}$ over a fixed $\mathbf{B}$ converges. In the CLIP prompt setting, the embedding-side coefficients carry the adaptation while the token basis can simply be fixed.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-10 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.09462 — 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 2026-10-16T10:02:34.301Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-10?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).