Search papers, labs, and topics across Lattice.
SynWeaver introduces a novel framework for web agents that synthesizes tasks and trajectories by leveraging structured website exploration to create a comprehensive website map. This approach addresses the limitations of existing methods by providing website-specific supervision, which enhances the diversity and efficiency of task proposals and execution trajectories. Experimental results on WebArena and WebVoyager show that SynWeaver significantly outperforms traditional synthesis baselines, improving generalization capabilities for web agents on both familiar and unfamiliar websites.
By integrating structured website exploration with task-trajectory synthesis, SynWeaver enables web agents to achieve unprecedented levels of generalization across diverse websites.
Web agents often struggle to generalize to unseen websites because they lack website-specific supervision. Recent exploration-based data synthesis methods reduce manual annotation, but they still face two key limitations: they often fail to cover the full functionality of a website, and without sufficient website prior knowledge, they tend to propose hallucinated tasks, which in turn limits the diversity and efficiency of downstream trajectory synthesis. We present \textbf{SynWeaver}, a website-prior task-trajectory co-synthesis framework designed to address these challenges. SynWeaver first performs structured website exploration and constructs a website map that covers a broad set of functionally distinct page states and executable interactions on the target website. It then derives page-level and transition-level supervision from this map to train a UI-aware model with website-specific priors, enabling more grounded task proposals. Finally, SynWeaver performs collaborative task-trajectory synthesis, jointly updating the task and execution trajectory when they become inconsistent, and then verifies and repairs the collected results to produce executable, semantically aligned supervision. Experiments on WebArena and WebVoyager demonstrate that SynWeaver consistently outperforms strong synthesis baselines and yields more effective supervision for both in-domain and out-of-domain generalization.
