Event ·Multi-source

Agent Arena 分析 20840 条轨迹:编码智能体差评首因是代码跑不通

Arena 团队分析了 Agent Arena: Code 平台上 29 个模型的 20840 条用户反馈轨迹,用直接的正负评价来衡量编码智能体的实际表现。投诉中 69.1% 指向代码无法运行,其后是输出不完整(27.1%)、收尾与可用性差(27.1%)和忽视指令(20.3%)。Fable 5.1 的“slop”与设计类投诉占 3.9%,低于 Astra 的 5.7%,“flaky”类投诉占 3.1%,也低于 Astra 的 5.0%。29 个模型的数据还显示,各实验室的新一代前沿模型获得的正反馈占比普遍更高。

Current read

Arena 拆了 20840 条编码智能体反馈:69% 差评是代码跑不通,Fable 5.1 比 Astra 更少被吐槽。

11 sources58° AI heatLast updated 2026/9/21 15:38:58

Evidence

11 sources
Related source 3eric zakariasson
View source
Related source 4shao__meng
View source
Related source 6向阳乔木
View source
Related source 7歸藏(guizang.ai)
View source
Related source 8arXiv: DeepSeek
View source
Related source 9Browser Use
View source
Related source 10OpenAI Blog
View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.