Metraipolis › archive › dc_c81b2000e57e

AcFlow: Controlling Text-to-Image Diffusion Transformers via Learned Conditional Activation Flow — arXiv:2609.10723

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.10723
License (as archived): CC0-1.0
Archived: 2026-10-16T05:24:36.377Z
SHA-256 of the archived text: ac7c4c74e0e35baa79d1b3e97cb2ac58bc439f3f2cea982da230561b26ece94b
Stable URL of this page: https://metraipolis.com/documento/dc_c81b2000e57e

How to cite

AcFlow: Controlling Text-to-Image Diffusion Transformers via Learned Conditional Activation Flow — arXiv:2609.10723 — arXiv (en), https://arxiv.org/abs/2609.10723 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c81b2000e57e (Metraipolis archive dc_c81b2000e57e, sha256 ac7c4c74e0e35baa79d1b3e97cb2ac58bc439f3f2cea982da230561b26ece94b)

BibTeX

@misc{dcc81b2000e57e,
  title = {AcFlow: Controlling Text-to-Image Diffusion Transformers via Learned Conditional Activation Flow — arXiv:2609.10723},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.10723},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c81b2000e57e; sha256 ac7c4c74e0e35baa79d1b3e97cb2ac58bc439f3f2cea982da230561b26ece94b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.10723. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Text-to-image diffusion transformers (DiTs) are powerful generators, yet direct prompting provides limited control interface for style intensity and can fail to suppress unwanted concepts. To enable these controls, we introduce AcFlow, an inference-time controller that transports intermediate layer image-token activations through a learned concept-conditioned velocity field while keeping the base DiT frozen. A textual concept description specifies the desired intervention, while the integration horizon provides a continuous control parameter. The field produces token-varying, activation-dependent updates. With parameters shared across concepts within each task family, one field covers over 15,000 style descriptions or over 1,000 suppression concepts, and generalizes to concepts unseen during training without per-concept fitting. On style control, AcFlow achieves the best style--content trade-off among the evaluated baselines in the high-style-alignment regime. At a fixed operating point, AcFlow attains style--content alignment of 0.5365/0.2860, compared with 0.4397/0.2684 for the baseline with the highest style alignment. On concept suppression, AcFlow reduces the fraction of images showing the concept from 95.3%/82.1% to 41.6%/40.5% on held-in/held-out concepts, including cases where deleting them from the prompt fails to remove them. Our analyses support the learned velocity field as an adaptive control mechanism, with update directions varying across tokens and depending on their activation states. Our code is available at this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-11 in cs.AI).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.10723 — 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 2026-10-16T05:17:52.072Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-11?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).