Metraipolis › archive › dc_9bd85d90100d

Decompose the Distillation: Interpretable Single-Pass Guidance for Diffusion Models — arXiv:2611.05755

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.05755
License (as archived)
CC0-1.0
Archived
2026-11-05T18:29:45.223Z
SHA-256 of the archived text
dd52f544bd698eabcb78be3f7c41623b0e341744af58e341c15ee383eb6f03f1
Stable URL of this page
https://metraipolis.com/documento/dc_9bd85d90100d

How to cite

Decompose the Distillation: Interpretable Single-Pass Guidance for Diffusion Models — arXiv:2611.05755 — arXiv (en), https://arxiv.org/abs/2611.05755 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_9bd85d90100d (Metraipolis archive dc_9bd85d90100d, sha256 dd52f544bd698eabcb78be3f7c41623b0e341744af58e341c15ee383eb6f03f1)
BibTeX
@misc{dc9bd85d90100d,
  title = {Decompose the Distillation: Interpretable Single-Pass Guidance for Diffusion Models — arXiv:2611.05755},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.05755},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_9bd85d90100d; sha256 dd52f544bd698eabcb78be3f7c41623b0e341744af58e341c15ee383eb6f03f1},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.05755. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Classifier-Free Guidance is a widely used inference-time guidance rule for conditional image generation with diffusion models, which exposes a guidance scale that adjusts conditioning strength after training. This flexibility requires conditional and unconditional model evaluations at every sampling step, so an N-step trajectory takes 2N forward passes. Single-pass alternatives reduce such cost, yet most collapse CFG's two components into one opaque guided prediction, losing the separable base flow and guidance direction that support per-step control and diagnostics, even when the scale remains adjustable. This work introduces CtrlFlow, a single-pass guidance method that preserves CFG's separable base-flow and guidance-direction interface. To preserve both components at one-pass cost, CtrlFlow attaches two lightweight LoRA adapters to a frozen backbone, each adding 2.5% parameters, and predicts the base flow and guidance direction from one pass of the shared frozen backbone. Because the guidance scale multiplies the guidance direction only after the forward pass, one evaluation yields the guided velocity at any scale, so a single trained checkpoint remains adjustable at inference without retraining. Preserving two outputs introduces a training mismatch: inference amplifies guidance-direction errors by the chosen scale, while naive component-wise losses weight both outputs equally. To align training with this composed inference-time error, CtrlFlow uses a unified multi-scale distillation loss. Residual guidance errors can still persist at extreme scales, where even small guidance-direction errors are amplified by the guidance scale. To reduce this inference-time error, CtrlFlow applies scale-adaptive modulation to damp velocity-field errors and time-grid warping to concentrate solver evaluations where the trajectory changes most rapidly.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Graphics (cs.GR).
Pagina del lavoro: https://arxiv.org/abs/2611.05755 — 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 2026-11-05T18:19:49.040Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).