arXiv cs.AI· Haldun Balim, Yilun Du, Na Li·· 3 天前AI 评分31
GTP：基于生成式轨迹规划的模型在线决策框架
Model-Based Online Decision Making via Generative Trajectory Planning
AI 导读
研究者提出 Generative Trajectory Planning（GTP），一种基于模型的框架，通过从扩散先验中采样候选动作序列并利用学习到的动力学、奖励和价值模型进行精炼，实现高效的轨迹级推理。在连续控制和长时程操作任务中，GTP 匹配或超越 SOTA 的无模型、扩散策略和基于模型的基线，并在强基线退化的场景下保持稳定。
正文 · 原文
Abstract:Online decision-making faces a tension between efficient inference and long-horizon reasoning. Value-based methods rely on short-horizon bootstrapping that limits the propagation of long-term information, while trajectory optimization methods plan explicitly but must solve a high-dimensional optimization problem at every decision step. We propose Generative Trajectory Planning (GTP), a model-based framework that addresses this tension by learning a generative model as a proposal distribution over action sequences. Instead of optimizing sequences from scratch, GTP samples candidate sequences from a diffusion prior and refines them using learned dynamics, reward, and value models, enabling efficient trajectory-level reasoning. Across continuous control and long-horizon manipulation tasks, GTP matches or exceeds state-of-the-art model-free, diffusion-policy, and model-based baselines, and remains stable in regimes where strong baselines degrade. Our results suggest that generative models are best understood not as policies, but as structured proposal distributions for planning, offering a new perspective on how to integrate generative modeling with model-based decision-making.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Robotics (cs.RO); Systems and Control (eess.SY) |
| Cite as: | arXiv:2610.18677 [cs.LG] |
| (or arXiv:2610.18677v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.18677 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Haldun Balim [view email]
[v1]
Tue, 13 Oct 2026 16:35:13 UTC (1,864 KB)
来源：arXiv cs.AI · arxiv.org