webexplorer-long-horizon-web-agent-training

【开发与工程】Systematically train 8B web agents to solve complex information-seeking tasks through model-based exploration and long-to-short query evolution, achieving state-of-the-art performance via SFT+RL training pipeline with progressive context expansion to 128K tokens and 100 tool turns.

独立 GitHub 快照 · 自动分类 · 以原作者文档为准