Dongxi 东锡 NLP@dongxi_nlp第31周论文 探索式建模——模型探索多条有效路径，每种可能性都保持清晰。 StateAct——智能体先读取应用状态，像素沦为后备方案，动作更加有据可依。 论文： 探索式建模：解锁第三条预训练轴与端到端生成 StateAct：程序状态优先于像素，用于长时程计算机操作智能体
Papers of Week 31
Explorative Modeling The model explores many valid paths. Each possibility stays sharp.
StateAct Agents read app state first. Pixels become a fallback. Actions become more grounded.
Papers：
Explorative Modeling： Unlocking a Third Pretraining Axis and End-to-End Generation
StateAct： Program State， before Pixels， for Long-Horizon Computer-Use Agents
第31周论文 探索式建模——模型探索多条有效路径，每种可能性都保持清晰。 StateAct——智能体先读取应用状态，像素沦为后备方案，动作更加有据可依。 论文： 探索式建模：解锁第三条预训练轴与端到端生成 StateAct：程序状态优先于像素，用于长时程计算机操作智能体
Papers of Week 31
Explorative Modeling The model explores many valid paths. Each possibility stays sharp.
StateAct Agents read app state first. Pixels become a fallback. Actions become more grounded.
Papers：
Explorative Modeling： Unlocking a Third Pretraining Axis and End-to-End Generation
StateAct： Program State， before Pixels， for Long-Horizon Computer-Use Agents