SCOPE 框架：让 LLM Web 智能体从局部技能学成长程任务能力
From Local Skills to Long-Horizon Web Tasks: Progressive Skill Exploration for LLM Web Agents
训练 Web 智能体老是拿不到反馈？这篇把复杂任务拆成技能链一步步学的思路挺有意思，三个基准上最多提升 18.6%。
一篇 arXiv 论文提出 SCOPE（Skill-Composed Progressive Exploration），借鉴技能习得理论训练长程 Web 智能体。该方法包含三个环节：Skill Discovery 从历史 rollout 中提取可复用局部技能与失败经验，Skill Orchestration 把技能组合成高层技能链，Skill Expansion 利用失败经验引导探索新子技能。通过 agentic on-policy distillation 将技能链内化到策略参数。在 WebArena、WebVoyager、WebShop 三个基准上，SCOPE 相比各设置下最强基线取得 3.9% 到 18.6% 的相对提升。
From Local Skills to Long-Horizon Web Tasks: Progressive Skill Exploration for LLM Web Agents
Large language models (LLMs) are driving autonomous agents toward real-world environments, with the Web emerging as one of the most representative interaction scenarios. However, training Web agents with naive exploration remains challenging: successful trajectories are rare, and the sparse terminal rewards make it difficult for agents to obtain useful feedback from failed trajectories. 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. Specifically, SCOPE iteratively explores around local skills: Skill Discovery extracts reusable local skills and failure experiences from historical rollouts; Skill Orchestration composes discovered skills into high-level skill chains; and Skill Expansion leverages failure experiences aligned with the current skill chain to guide the agent toward exploring new subskills. Through this process, SCOPE progressively constructs task-completing skill chains across rollouts and internalizes them into the policy parameters through agentic on-policy distillation. Extensive experiments on WebArena, WebVoyager, and WebShop show that SCOPE consistently improves exploration dynamics and achieves 3.9% to 18.6% relative gains over the strongest baselines in each setting.
