Metraipolis › archive › dc_270a03495445

Beyond Future Prediction: Denoising as Generative Adaptation for Robot Control — arXiv:2609.28339

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.28339
License (as archived): CC0-1.0
Archived: 2026-10-08T12:53:52.903Z
SHA-256 of the archived text: 1173ee8892249dfc9ecdcb02832e4909f46cbb7b2874d06abfd384056a5f7c65
Stable URL of this page: https://metraipolis.com/documento/dc_270a03495445

How to cite

Beyond Future Prediction: Denoising as Generative Adaptation for Robot Control — arXiv:2609.28339 — arXiv (en), https://arxiv.org/abs/2609.28339 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_270a03495445 (Metraipolis archive dc_270a03495445, sha256 1173ee8892249dfc9ecdcb02832e4909f46cbb7b2874d06abfd384056a5f7c65)

BibTeX

@misc{dc270a03495445,
  title = {Beyond Future Prediction: Denoising as Generative Adaptation for Robot Control — arXiv:2609.28339},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.28339},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_270a03495445; sha256 1173ee8892249dfc9ecdcb02832e4909f46cbb7b2874d06abfd384056a5f7c65},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.28339. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Pretrained generative Diffusion Transformers (DiTs) capture rich pixel-level visual and language-conditioned structure through large-scale image and video generation training. A growing line of robot policies builds on this generative prior, but how it should be transferred to control remains unclear, and existing approaches commonly instantiate this transfer through future visual prediction. We ask a more basic question: what a pretrained generative DiT actually contributes to action learning, and how this prior should be adapted for control. We introduce NowWAM, a future-target-free co-training formulation that denoises the current observation and predicts robot actions from the same visual stream, directly coupling the native generative objective to the action-facing representation across the denoising trajectory. Under matched controlled settings, past and future visual targets perform comparably, while restricting training to the clean endpoint substantially reduces robustness, suggesting that a separate future target is not essential for generative adaptation, while the denoising trajectory remains an effective interface for control. On LIBERO-Plus, NowWAM reaches 87.7% with FLUX2-Klein, improving over the future-target co-training baseline by 6.1 points while halving training visual tokens (784 to 392) and reducing step time from 2.85 s to 1.63 s, a 1.8x speedup. With the pure text-to-image Z-Image backbone, NowWAM further reaches 87.8%, showing that strong control adaptation is not tied to video generation or image-editing backbones.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-24 in cs.RO).
Categorie arXiv: Robotics (cs.RO).
Pagina del lavoro: https://arxiv.org/abs/2609.28339 — 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 2026-10-08T12:21:19.513Z dalla pagina https://arxiv.org/catchup/cs.RO/2026-09-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).