Rohan Paul· @rohanpaul_ai · X·· 2 小时前AI 评分34
AI 导读
– https://arxiv.org/abs/2610.11050 标题:"AgentHorizon:为长周期计算机使用任务评估智能体评判器"
正文
– https://arxiv.org/abs/2610.11050
Title: "AgentHorizon: Evaluating Agentic Judges for Long-Horizon Computer-Use Tasks"
来源:Rohan Paul · x.com