arXiv · cs.AI · Xueyang Feng, Xiaohe Bo, Xu Chen, Quanyu Dai, Feng Wen · 10 小时前模型评分43
SCOPE:让 大语言模型(LLM,Large Language Model) 网页智能体从局部技能渐进探索长程任务
From Local Skills to Long-Horizon Web Tasks: Progressive Skill Exploration for LLM Web Agents
导读
研究者提出面向长程网页任务的短到长学习框架 Skill-Composed Progressive Exploration(SCOPE),把复杂任务探索拆解为从局部技能到完整任务能力的渐进扩展过程。
分类置信度
中 · 待人工复核
- 只有单一媒体来源,需要人工复核。
来源级别:媒体与个人
事件类型:研究发布
证据摘录
- “Inspired by skill acquisition theory, we propose Skill-Composed Progressive Exploration (SCOPE), a short-to-long learning framework for long-horizon Web agents that decomposes complex task exploration into a gradual expansion process from local skills to complete task-solving capability.”
来源:arXiv · cs.AI · arxiv.org