Boosting Deepresearch and LongContext Ability with Self-Generated Deepresearch Rollouts Traces
The method repurposes DR-RL trajectories, which naturally contain search histories, visited webpages, evidence snippets, and final-answer supervision, and replaces the compact snippets and webpage summaries in each trajectory with the full contents of their corresponding URLs, producing substantially longer multi-docum...
Zi-Han Wang, Hao Wang, Bo Jiang et al.
· 0 citations