AI News HubLIVE
原文2 min read

QPILOTS: Efficient Test-Time Q-Steering for Flow Policies

QPILOTS is a method that steers the denoising process at inference time using Q-values without modifying the original policy. It projects intermediate noisy states to an estimate of the final clean action to compute critic gradients, avoiding instability from direct backpropagation. It achieves 90% average success rate across 50 offline-to-online RL tasks and outperforms prior inference-time approaches on simulated manipulation tasks using a VLA model.

SourcearXiv Machine LearningAuthor: Yifan Ruan, Chenyang Cao, Andreas Burger, Ali Pesaranghader, Kaveh Kamali, Jaehong Kim, Nandita Vijaykumar, Alan Aspuru-Guzik, Igor Gilitschenski, Nicholas Rhinehart

[2606.14801] QPILOTS: Efficient Test-Time Q-Steering for Flow Policies

[Submitted on 11 Jun 2026]

Title:QPILOTS: Efficient Test-Time Q-Steering for Flow Policies

View a PDF of the paper titled QPILOTS: Efficient Test-Time Q-Steering for Flow Policies, by Yifan Ruan and 9 other authors

View PDF HTML (experimental)

Abstract:Flow-matching and diffusion policies are expressive action generators, but optimizing them with temporal-difference reinforcement learning (RL) remains difficult. Effective policy extraction requires exploiting the critic's action gradient, yet directly backpropagating this signal through a multi-step denoising process can be numerically unstable. Existing methods work around this either by discarding gradient information, distilling the policy into a simpler one-step actor, or repeatedly fine-tuning the denoising policy as the critic improves. We propose QPILOTS, a method that leaves the original policy unmodified and steers the denoising process at inference time. At each denoising step, instead of evaluating the critic on the noisy intermediate action where critic predictions are unreliable, we first project that intermediate state to an estimate of the final clean action and compute the critic gradient there. We introduce two variants: QPILOTS-U uses a fast single-point approximation, while QPILOTS-M draws differentiable posterior samples via a learned auxiliary network. On a standard offline-to-online RL benchmark, QPILOTS achieves the best aggregate performance, reaching an average success rate of 90% across 50 tasks. We also apply QPILOTS to steer a large, frozen, pretrained Vision-Language Action (VLA) foundation model, outperforming or matching prior inference-time approaches across six manipulation tasks in simulation.

Comments: 10 pages, 7 figures

Subjects:

Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Robotics (cs.RO)

Cite as: arXiv:2606.14801 [cs.LG]

(or arXiv:2606.14801v1 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2606.14801

arXiv-issued DOI via DataCite

Submission history

From: Yifan Ruan [view email] [v1] Thu, 11 Jun 2026 18:22:03 UTC (3,441 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled QPILOTS: Efficient Test-Time Q-Steering for Flow Policies, by Yifan Ruan and 9 other authors

View PDF

HTML (experimental)

TeX Source

view license

Current browse context:

cs.LG

new | recent | 2026-06

Change to browse by:

cs cs.AI cs.RO

References & Citations

NASA ADS

Google Scholar

Semantic Scholar

Loading...

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

IArxiv recommender toggle

IArxiv Recommender (What is IArxiv?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)