跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.10343 · v1

GUARD: Anomaly-Informed Q-Steering for Generative Robot Policies in Dexterous Manipulation

首次提交 2026-11-07 08:26
版本更新 2026-11-07 08:26
雷达观测 2026-11-10 10:11
arXiv ↗PDF ↗

摘要

Generative robot policies based on diffusion and flow matching can model complex, multimodal behaviors, but remain susceptible to covariate shift during closed-loop execution. Existing approaches mitigate such failures through additional demonstrations, policy fine-tuning, or test-time guidance, but often require human intervention, engineered rewards, or other task-specific signals. We introduce GUARD, an anomaly-informed Q-steering framework that leverages the policy's training distribution to construct a dense, failure-anticipatory steering signal. GUARD pairs sparse task rewards with dense anomaly scores learned from policy training data and propagates both through a learned Q-function. Its gradients then steer the denoising process toward actions that improve task return while preventing trajectories from entering poorly supported regions, without directly modifying the pretrained policy. Across four Adroit dexterous manipulation tasks, GUARD achieves average success rates of 96.0%96.0\% and 80.0%80.0\% under nominal and perturbed dynamics, outperforming the strongest baseline by 21.021.0 and 20.020.0 percentage points while maintaining the lowest anomaly score. GUARD further demonstrates direct sim-to-real transfer on in-hand screwdriver rotation and reorientation tasks, achieving 4040 and 2525 percentage-point gains in success rate. These results demonstrate that anomaly signals can serve not only for failure monitoring but also to complement sparse task rewards as effective supervision for robust policy steering.

作者备注:8 pages, 4 figures

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

目前只收录了一个版本。收录下一版后,可以对照标题、作者、分类和摘要。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v1 · 2026-11-07 08:26 · 最新观测版本

GUARD: Anomaly-Informed Q-Steering for Generative Robot Policies in Dexterous Manipulation

Generative robot policies based on diffusion and flow matching can model complex, multimodal behaviors, but remain susceptible to covariate shift during closed-loop execution. Existing approaches mitigate such failures through additional demonstrations, policy fine-tuning, or test-time guidance, but often require human intervention, engineered rewards, or other task-specific signals. We introduce GUARD, an anomaly-informed Q-steering framework that leverages the policy's training distribution to construct a dense, failure-anticipatory steering signal. GUARD pairs sparse task rewards with dense anomaly scores learned from policy training data and propagates both through a learned Q-function. Its gradients then steer the denoising process toward actions that improve task return while preventing trajectories from entering poorly supported regions, without directly modifying the pretrained policy. Across four Adroit dexterous manipulation tasks, GUARD achieves average success rates of 96.0%96.0\% and 80.0%80.0\% under nominal and perturbed dynamics, outperforming the strongest baseline by 21.021.0 and 20.020.0 percentage points while maintaining the lowest anomaly score. GUARD further demonstrates direct sim-to-real transfer on in-hand screwdriver rotation and reorientation tasks, achieving 4040 and 2525 percentage-point gains in success rate. These results demonstrate that anomaly signals can serve not only for failure monitoring but also to complement sparse task rewards as effective supervision for robust policy steering.

查看此版本解读 →查看此版本原文 ↗