AI 导读
– https://t.co/qJoaEvvag3 标题："Long-Horizon-Terminal-Bench：用基于密集奖励的评分测试智能体在长时程终端任务上的极限"
正文
– https://t.co/qJoaEvvag3
Title: "Long-Horizon-Terminal-Bench: Testing the Limits of Agents on Long-Horizon Terminal Tasks with Dense Reward-Based Grading"
来源：@rohanpaul_ai · x.com