Metraipolis › archive › dc_35afdb424ea3

Safe Real-Time Policy Steering via Noise-Space Trajectory Optimization for One-Step Generative Policies — arXiv:2609.21220

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.21220
License (as archived): CC0-1.0
Archived: 2026-10-10T02:30:03.440Z
SHA-256 of the archived text: c1abfc6ce28c79087c52e22c3e735efd31016f5d9c430a780ec07e7b6dbc1d9e
Stable URL of this page: https://metraipolis.com/documento/dc_35afdb424ea3

How to cite

Safe Real-Time Policy Steering via Noise-Space Trajectory Optimization for One-Step Generative Policies — arXiv:2609.21220 — arXiv (en), https://arxiv.org/abs/2609.21220 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_35afdb424ea3 (Metraipolis archive dc_35afdb424ea3, sha256 c1abfc6ce28c79087c52e22c3e735efd31016f5d9c430a780ec07e7b6dbc1d9e)

BibTeX

@misc{dc35afdb424ea3,
  title = {Safe Real-Time Policy Steering via Noise-Space Trajectory Optimization for One-Step Generative Policies — arXiv:2609.21220},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.21220},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_35afdb424ea3; sha256 c1abfc6ce28c79087c52e22c3e735efd31016f5d9c430a780ec07e7b6dbc1d9e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.21220. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Generative robot policies can represent diverse, multimodal behaviors, but adapting pretrained policies to deployment-time constraints such as collision avoidance and orientation maintenance remains challenging. Existing inference-time steering methods typically apply gradient guidance through iterative diffusion or flow processes, which can be computationally expensive for real-time control. We propose INSPO, which formulates inference-time steering of one-step generative policies as trajectory optimization in the policy's input noise space. By optimizing the input noise while evaluating constraints on the induced state trajectory, INSPO searches the policy-induced behavior space without directly modifying generated actions. The optimization includes a regularization term that encourages solutions to remain consistent with the policy's input distribution and is solved online using population-based particle optimization. We evaluate INSPO on state- and image-based task-specific policies and generalist vision-language-action policies across Push-T, Can pick-and-place, and LIBERO-Spatial. INSPO improves task success and constraint satisfaction over best-of-N sampling and action projection, while comparing favorably with gradient-guided generation at lower runtime.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-21 in cs.RO).
Categorie arXiv: Robotics (cs.RO).
Pagina del lavoro: https://arxiv.org/abs/2609.21220 — 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 2026-10-10T02:09:55.322Z dalla pagina https://arxiv.org/catchup/cs.RO/2026-09-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).