智能体
arxiv.org ·
StraTA：通过策略轨迹抽象激励智能体强化学习
该论文提出 StraTA（Strategic Trajectory Abstraction）框架，在智能体强化学习中引入显式的轨迹级策略：从初始任务状态采样一个紧凑策略，并让后续动作以该策略为条件，通过分层 GRPO 式 rollout 联合训练策略生成与动作执行，并加入多样化策略 rollout 与关键自评判。实验在 ALFWorld、WebShop、SciWorld 上进行，报告 ALFWorld 成功率 93.1%、WebShop 84.2%、SciWorld 总分 63.5%，作者称在样本效率与最终性能上均优于强基线，并在 SciWorld 上超过前沿闭源模型。
AI推荐理由 arXiv 正式论文（v2，2026-09-25 修订），作者含 Philip Torr、Wanli Ouyang、Lei Bai 等，摘要给出明确方法与量化结果：StraTA 在初始任务状态采样紧凑策略并以此条件化后续动作，用分层 GRPO 式 rollout 联合训练策略生成与动作执行，辅以多样化策略 rollout 与关键自评判；ALFWorld 成功率 93.1%、WebShop 84.2%、SciWorld 总分 63.5%，称优于前沿闭源模型。新信息在于把显式轨迹级策略引入 agentic RL，针对长程决策中纯反应式方法探索与信用分配弱的问题。局限：仅摘要级证据，未提供基线细节、消融与统计显著性，'优于前沿闭源模型'的具体对比对象未列出，需查全文核实。
78
正文
Computer Science > Computation and Language
arXiv:2605.06642 (cs)
[Submitted on 7 May 2026 (v1), last revised 25 Sep 2026 (this version, v2)]
Title:StraTA: Incentivizing Agentic Reinforcement Learning with Strategic Trajectory Abstraction
Authors:Xiangyuan Xue, Yifan Zhou, Zidong Wang, Shengji Tang, Philip Torr, Wanli Ouyang, Lei Bai, Zhenfei Yin
View a PDF of the paper titled StraTA: Incentivizing Agentic Reinforcement Learning with Strategic Trajectory Abstraction, by Xiangyuan Xue and 7 other authors
View PDF
HTML (experimental)
Abstract:Large language models (LLMs) are increasingly used as interactive agents, but optimizing them for long-horizon decision making remains difficult because current methods are largely purely reactive, which weakens both exploration and credit assignment over extended trajectories. In this work, we present Strategic Trajectory Abstraction (StraTA), a simple framework that introduces an explicit trajectory-level strategy into agentic reinforcement learning (RL). StraTA samples a compact strategy from the initial task state, conditions subsequent actions on that strategy, and trains strategy generation and action execution jointly with a hierarchical GRPO-style rollout design, further enhanced by diverse strategy rollout and critical self-judgment. Experiments on ALFWorld, WebShop, and SciWorld show that StraTA consistently improves both sample efficiency and final performance over strong baselines. StraTA reaches success rates of 93.1% on ALFWorld and 84.2% on WebShop. On SciWorld, StraTA attains a 63.5% overall score, outperforming frontier closed-source models.
Subjects:
Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
Cite as:
arXiv:2605.06642 [cs.CL]
(or
arXiv:2605.06642v2 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2605.06642
Focus to learn more
arXiv-issued DOI via DataCite
Submission history
From: Xiangyuan Xue [view email]
[v1]
Thu, 7 May 2026 17:51:16 UTC (1,975 KB)
[v2]
Fri, 25 Sep 2026 17:48:02 UTC (1,649 KB)
Full-text links:
Access Paper:
View a PDF of the paper titled StraTA: Incentivizing Agentic Reinforcement Learning with Strategic Trajectory Abstraction, by Xiangyuan Xue and 7 other authors
View PDF
HTML (experimental)
TeX Source
view license
Current browse context:
cs.CL
new
|
recent
| 2026-05
Change to browse by:
cs
cs.AI
References & Citations
NASA ADS
Google Scholar
Semantic Scholar
export BibTeX citation
Loading...
BibTeX formatted citation
×
loading...
Data provided by:
Bookmark
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?)
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?)