Btw, I think GLM-5.1 was trying to do something very ambitious here, and failed due to fumbling step size
令人惊讶的是：GLM-5.1作为一个先进AI模型，竟然因为'步长处理不当'这种技术细节而失败，这表明即使是顶级AI也可能在基础执行层面出现问题，而不仅仅是概念设计上的不足。
4 Matching Annotations
- Last 7 days
-
-
-
rdi.berkeley.edu rdi.berkeley.edu
-
We built an automated scanning agent that systematically audited eight among the most prominent AI agent benchmarks — SWE-bench, WebArena, OSWorld, GAIA, Terminal-Bench, FieldWorkArena, and CAR-bench — and discovered that every single one can be exploited to achieve near-perfect scores without solving a single task.
令人惊讶的是：研究人员构建的自动化扫描工具发现，所有八个主流AI代理基准测试都存在漏洞，无需解决任何任务就能获得接近完美的分数。这表明整个AI评估领域存在系统性问题，几乎所有当前使用的基准测试都不可靠。
-
- Mar 2026
-
ianarawjo.com ianarawjo.com
-
Decades spent educating researchers have had little or no influence on beliefs and practice (Schmidt and Hunter, 1997, pp.20–22).
Calls for reform fall on deaf ears
-
NHST has been severely criticized for more than 50 years by end users to whom fair statistical communication matters.
Calls for reform fall on deaf ears
-