English(EN) SAVGO: Learning State-Action Value Geometry with Cosine Similarity for Continuous Control

SAVGO算法利用几何学改进强化学习策略更新

作者 PulseAugur 编辑部 · [2 个来源] · 2026-05-01 17:09

研究人员推出了一种新颖的强化学习算法SAVGO，旨在改进连续控制任务中的策略更新。SAVGO学习一个联合状态-动作嵌入空间，其中相似的动作-值估计由高余弦相似度表示。这种几何方法允许策略改进朝着更高价值区域引导，统一了表示学习、值估计和策略优化。在MuJoCo基准测试上的评估表明，SAVGO在复杂、高维任务上的表现优于现有方法。 AI

影响在连续控制强化学习中引入了一种新的策略更新几何方法，有望提高复杂任务的样本效率和性能。

排序理由详细介绍一种新的强化学习算法的学术论文。

在 arXiv cs.LG 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。我们如何撰写摘要 →

报道来源 [2]

arXiv cs.LG TIER_1 English(EN) · Stavros Orfanoudakis, Pedro P. Vergara · 2026-05-04 04:00

SAVGO: Learning State-Action Value Geometry with Cosine Similarity for Continuous Control

arXiv:2605.00787v1 Announce Type: new Abstract: While representation and similarity learning have improved the sample efficiency of Reinforcement Learning (RL), they are rarely used to shape policy updates directly in the action space. To bridge this gap, a geometry-aware RL algo…
arXiv cs.LG TIER_1 English(EN) · Pedro P. Vergara · 2026-05-01 17:09

SAVGO: Learning State-Action Value Geometry with Cosine Similarity for Continuous Control

While representation and similarity learning have improved the sample efficiency of Reinforcement Learning (RL), they are rarely used to shape policy updates directly in the action space. To bridge this gap, a geometry-aware RL algorithm that explicitly incorporates value-based s…

报道来源 [2]

SAVGO: Learning State-Action Value Geometry with Cosine Similarity for Continuous Control

SAVGO: Learning State-Action Value Geometry with Cosine Similarity for Continuous Control

相关实体

相关话题