arXiv cs.AI· Minkyoung Cho, Zewei Zhou, Wenhao Ding, Shuhan Tan, Boyi Li, Yuxiao Chen, Yan Wang, Zheng Lian, Min-Hung Chen, Chaowei Xiao, Zhuoqing Mao, Boris Ivanovic, Marco Pavone, Yulong Cao·· 10 天前AI 评分22
GroundAct：面向自动驾驶物理智能的 grounded planning 新框架
Grounding What Shapes the Plan: Rethinking Groundedness for Physical Intelligence in Autonomous Driving
AI 导读
研究团队提出 GroundAct，将物理实体作为 grounding 单元，用轻量 reference token 保持每个被选实体连续状态可寻址，仅由被引用实体与演化方案的交互来修正规划，形成从推理依据到规划动作的显式路径。在开环与闭环评测中，GroundAct 在正常、分布外及安全关键场景下均展现出较强的规划能力，闭环结果进一步在仿真驾驶中得到验证。
正文 · 原文
Authors:Minkyoung Cho, Zewei Zhou, Wenhao Ding, Shuhan Tan, Boyi Li, Yuxiao Chen, Yan Wang, Zheng Lian, Min-Hung Chen, Chaowei Xiao, Zhuoqing Mao, Boris Ivanovic, Marco Pavone, Yulong Cao
Abstract:Driving models increasingly ground reasoning in causal relations, spatial structure, perceptual evidence, and predicted futures. These advances make reasoning more faithful to the driving scene, but leave a fundamental question unresolved: what should groundedness mean when the model ultimately outputs an action? Correctly grounded reasoning does not, by itself, ensure desirable driving outcomes. We introduce GroundAct, which starts from a simple premise: driving unfolds through physical entities and their interactions. Entities therefore become the unit of grounding; a lightweight reference token keeps each selected entity's continuous state addressable through symbolic reasoning; and only the referenced entities' interactions with the evolving proposal correct the plan. The result is an explicit path from what reasoning grounds to what the plan does, which we call grounded planning. To assess its practical value, we evaluate GroundAct in both open- and closed-loop settings. GroundAct shows strong open-loop planning across normal, out-of-distribution, and safety-critical scenarios, with closed-loop results extending this evidence to driving in simulation.
| Comments: | 20 pages; Project website: this https URL |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.07521 [cs.AI] |
| (or arXiv:2610.07521v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.07521 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Minkyoung Cho [view email]
[v1]
Mon, 5 Oct 2026 23:39:38 UTC (3,036 KB)
来源：arXiv cs.AI · arxiv.org